Brainstorm AI San Francisco 2024

Martin Casado, General Partner, Andreessen Horowitz, Dawn Song, Professor, Computer Science; Co-director, The Berkeley Center for Responsible, Decentralized Intelligence, University of California, Berkeley

Brainstorm AI 2024: How Much Should We Regulate AI

Conferences

Martin Casado, General Partner, Andreessen Horowitz, Dawn Song, Professor, Computer Science

Hi, folks. Thanks for being with us. Thanks for having us. Yeah, we figured we'd end with a nice lighthearted topic, you know. Uh, happy to help. Yeah. Oh, yeah, of course. Of course. Um, let me just level set here. Does a, I need to be regulated? Don? Oh, I think our answer is, is very similar, but I'll probably present a little bit different than that. You did, which is, um, a, I is a software system. Software system is regulated by rules that have been developed for the last 30 years. They're very robust. They've been part of an ongoing discourse. Um, and so in as much as software is regulated, A I already is regulated and then there's a question of, does it need new regulation? And the answer to that is we'd have to understand what is new about it, which is still a research question. And so until we understand that I would say if you regulated it, you may be regulating the wrong thing if you did it too early. And that could be, you know, even more negative than if you didn't regulate it. So, my view is we have existing robust regulatory regime, it is sufficient for now. And if we understand, you know, that A I is different in some way, we should, we should evolve that, but not until we understand that Dawn. Does that sound like responsible A I? Ok. So I can, I, a lot of that was cribbed by the way, from a lot of that was cribbed from a paper that Don wrote by the way. So I think so. Ok, so, so first let me summarize what is the current lay of land, please do. So if you ask, if you ask this question to the policy makers, they essentially already, you know, speak with their act. So I called this year, we have a certain proliferation of A I belt. So if you look at the federal level, there are about 100 and 20 a belts in flight. And if you look at the state level, there are at least 45 states that are, have A I built in flight about around 600 of them. So it's crazy. So I would say that if you ask this question, there's a madness about A I regulation. Um and going, going back to what my T said, I think there is really um a lot of nuance that we need to look at when we do A I regulation if we don't do it. Well, then for example, if we do it in a way which is mostly what current A I regulation is we have a lot of problems we can have. Um uh we can have a lot of potential unforeseen negative consequences and also even worse, we can actually have a lot of to prevent disasters outcomes. And also we can have fragmentation in community, which is what we have seen unfortunately through, you know, the recent process of some of the bills that will. Um so that's why actually recently we launched a new proposal with a number of leading A I researchers called on the Path for Science and evidence space A I policy, which is excited to address this question. Should we regulate A I or not? How do we even answer this question? So here, then the answer is very simple. It should be science and evidence based. We don't just jump out of the hat whether we should today, we think we should regulate A I tomorrow. We think no, we shouldn't, we should have science and evidence to speak for us. And so that's what the key of the proposal is about is that a policy should be informed by science and evidence and we should do science and evidence based approach for A I policy. And also I strongly believe that this is the best approach forward for us to develop the best A I policy and also the best approach forward to unite the community to reduce recommendation and lead us to a safer A I world. Now, we we are sitting right now in the State of California, we three are residents of the State of California. So let me start at the state level and then we can talk about the federal level if you don't mind. Um, we obviously had SB 1047 come and go as a, a candidly, a flashpoint for the industry was that science and evidence based. Uh, and by the way, that's hardly the only bill that went through uh uh the California. So that's a very good question. So that's actually the reason, I mean, now the reason, but that's one of the motivation is from what we have seen in the process again, like I said, uh so currently most of this A I regulation effort is quite a hoc you can talk about right, the different thresholds. But a lot of these things, we need better, more rigorous approach that gives us the science and evidence and also to really analyze. So, for example, in the proposal that I mentioned, we actually proposed five priorities for A I policy and we propose how you should actually analyze A I risk, it should be risk. I actually I wrote them down. Don it's understand risks, increase transparency, develop early warning mechanisms, develop technical mitigation and defense mechanisms and trust to reduce fragmentation. Those are the five of the proposal, right? Great. Thank you. So, so I think I see SB 1047 is a great example for us to talk about because you know, there are so many policies. So why is this the one that got people up in arms? Right? And so I'll tell you why, which is um again, we have 30 years of discourse on regulation and policy in tech, right? And so we kind of understand basic principles, we kind of understand the basic paradigms by which we apply them. And SB 1047 was a paradigmatic shift in the following way, which it actually included something called developer liability and what developer liability means, which is like if I'm building a piece of software and somebody else uses it and that causes some catastrophic event that I am liable now in like 40 years or 30 years of tech policy, we've never done this and, and maybe, maybe it's something worth doing with A I maybe, but it's a paradigmatic shift that is divergent from 30 to 40 years of policy. And so we're like, well, wait, you know, time out, like, listen, if you want to do policy fine, we kind of understand policy. This is a discourse that's been going on for a long time. You've got a lot of experts, but you're not listening to any of that, you're doing something entirely new and, and based on, you know, as don, who's a world expert on this would say no evidence. And so I would say, listen, it's a very important problem A I policy. It's a very important problem and like, listen, we can even do spurious policies if like we want to because that's what policymakers do and they're going to do spurious policies, like we're going to definitely see that happen. But, but please, let's not create a new paradigm in policy that we don't understand the adverse effects of without actually understanding the risks. And, and the reason is, is we're going to set a precedent and we may end up like policing entirely the wrong thing. And so again, there's a lot of like, there's a lot of, um, new laws and policy being proposed all of the time, but certain ones of them are, are, are particularly pernicious and this was one of them for exactly this reason. And so it, you know, Governor Newsom vetoed it. So it's, it's, it's a moot point. But where do we go from here, Martin? I mean, do we think that we're actually going to take the big step back and do what you're saying? So, I, I think they are very legitimate efforts that are tangible and near term and like winding their way through the courts, right? Things like copyright, I think is incredibly legitimate and we should understand it. Things like child protections are incredibly gate and we should understand it. And so everybody agrees on this and there's consensus and so we should focus on those. So let's focus on the tangible ones that we know are real. A lot of the SB 147 like the proponents are X risk people, like they're like, oh A I is going to take over the world. So we need to put in these precautions that don't make any sense, right? And so I would say, like, listen, let's like, put aside the science fiction, let's put aside like this kind of existentialist stuff, let's focus on stuff that we know is kind of near term and real and impacts people's day to day lives and, and, and by and large, I think that's actually what is happening. I think that we've kind of regained some sense. It's been five years in this journey. Like we as far as we can tell like this A I stuff is actually pretty safe and like we're actually focusing on tangible practical near term regulation. Um OK, so again, coming back, actually, we uh again in the proposal, we made five priorities and uh and the number one actually is to understand A I risk. And actually in there, we actually propose that we do need to understand A I risk in a very comprehensive way. So there are all the risks that my men, but there are also longer term risks. So I think people need to, right? I think it's helpful for the society to do comprehensive study, to analyze, to understand their risk. By now, I think as a as I mentioned that we want to be science and evidence based. So the first step is that we need to better understand the science and we need to collect the evidence. And so I think it is important that we study a wide ranging different types of risks, but we need to understand them. And then the policy needs to be essentially need to be appropriate to the, to the level of understanding that, that we have. I, I think it's actually, um so a lot of people that have been following this don't know how remarkable it is that dawns up here. And so one thing about like these policy wars in the past like 20 years ago is like the forefront was academia and hobbyists and we had these robust debates. And what's been very remarkable about the A I one is like when SB 104 and seven came, like, like academia was largely silent or hadn't weighed in and so forth. And so I think another thing that's been fantastic about this first wave of policy. It is start of the conversation and was brought in stakeholders that really, I would say are far more neutral than say myself, I'm a venture capitalist, right? So clearly, I've got a bias, right? So I should be a voice, but I should not drive the conversation though. In that case, I actually was driving the conversation, right? So I think we're now at a great place where we've got people like Don who are a real expert who are like, hey, listen, like this is a very important problem like, and I think academia is engaged and I think that's actually where we should start moving these discussions to. And so the very concrete next step, let's all fund don this out for us pro pro I get it. So, so let's um, I, I do want to come to the audience for questions in a minute, but II, I want to make the jump up to the federal level. OK. We're gonna have a change in the White House very imminently. The, the outgoing president outlined an A I framework. The incoming one is certainly planning to do just that. How are we going to stick to our science and evidence based guns in this environment? Don. Um That's a very good question, I think especially in the changing environment. I think it's even more important that we focus on like the science, evidence based approach. And uh in fact, we are discussing at all levels um including the federal level and even including like the National Academy and, and uh so, so far, actually, this proposal has received a huge positive feedback. We uh it's only last maybe about two months that we already have volunteer contributors from about 200 different institutions actually um volunteer to help uh with this uh with this effort. And also, yeah, we have received like a lot of positive huge support from civil society from a lot of different, different entities and so on. So we really hope that we can bring this as a whole community society effort. I think it's hugely important. It sounds like a plan. All right, let's go to the audience. Nobody knows what's going to happen under Trump when it comes to like a policy. It's like, nobody knows. Right? Like, I don't know, you know, and so like, on one hand, I'm cautiously optimistic because like what happened under Biden was terrible. It was like he had this executive order that made no sense. And then, you know, it was like this kind of all this kind of crazy A I policy stuff on the other side. Like let's be honest, like SB 1047, which I think is probably the worst, like probably the worst bill, I would say the United States personally because of the paradigmatic shift was supported by Elon, right? And so anybody that thinks that they know it's going to happen, I don't think that they really understand the players, it could go either way. So I'm cautiously optimistic. I think it's good to have a change. I think where we're coming from was actually very bad when it came to A I policy, but no idea, you know what it will actually look like. All right. That's, yeah. Yeah. Yeah, we should ask an A I agent about it. All right, let's let's go to the audience for questions, comments about regulation. And policy. Oh, we've stunned them. Look at that. Yes. Right over here, sir. Your name and who you're with? Uh John Durkan, chief technologist with ID A Ireland. Uh So I'd be very interested from the panel's uh point of view, like particularly looking at the EU A I act that came out recently and the perspective that that has and we saw with GDPR how that kind of set regulation around the world, but in the context of the EU A I Act, your kind of feelings on that, has it been too much focused on regulation or is it because it's took a risk based approach? And it's kind of looking at the idea of setting kind of this idea of being more based on research to make the informed decisions as it goes through. So just interested in your view, I know my answer, but I'll let you go. Um I think, right, I mean, eu in general, they do take a lead a lot of these regulatory efforts. Um and now actually still they have this code of practice after it's ongoing. So I think a lot of details are actually still in the, in the progress to figure out exactly like what the company needs to do to comply, what are the recommendations and so on. And I think they are taking a first step and I think it's leading in the, I mean, certainly it's helping making progress in the space but of course, as I mentioned earlier, there's a lot that we actually really do need to study, to figure out very diplomatic. So I, I come from a very, very poor part of Spain actually. Like, I've got it on the DM, which is like the, the Appalachia of Spain. I love it. It's a great place. Um, you know, uh, we have Alcaraz as a tennis player. He's great. We also have a phenomenal founder who started actually a great company in Ja I um and the company is called Manic. His name is Javier and he's done all of this great work. And today as we speak, he's back in Murcia, you know, this, this, you know, it's great part of Spain but like not known for its tech founders and entrepreneurs. Um and he, and, and by the way, magnific is in the gen A I pixel space. So it's like images and videos and stuff like that and he cannot access Sora. So Sora is the video model that was just released by Open A I yesterday, right? And so like you tell me, but from my standpoint, like when you have like a a, you know, a great founder who can't even engage in his own area because you can't access to something that like a company that Open air release. I think that's a strictly bad. So I think that the eu is totally bundling it. I think it's putting itself backwards and I think it's got to change its tune or it's going to get in trouble. But how do you really feel that more questions, please? Yes, we have one way in the back. Your name and who you're with? Hi, I'm Alana Con, uh, with, uh, hacker one and Admission. I'm a former federal regulator. So I understand the instinct to say we don't need anything. We're already regulated. We've got this unlock. Um, but the federal regulation process, specifically the rule making process, it does take science and evidence into consideration because of course, there has to be notice and comment and you have to be able to provide that information which they take into consideration when they do those rules in the absence while we're off here saying we don't need anything. The Eu Singapore, everyone else is essentially regulating for us. So like privacy, we don't have anything here in the US. We've been fighting it for years and so the eu essentially runs our regulation. So is that the right approach from your perspective? And it doesn't make sense to continue to say we don't need anything and we'll let all these other countries do it for us. Great question. Thank you. Um So first, I don't think we are saying that we don't need anything right? I mean, the the proposal for a science and evidence based approach for a policy does not mean that we don't need anything. Actually, in fact, in the proposal, we actually talk about certain priorities for regulation, including for example, transparency, average event reporting. So we can write to like for post deployment harm and do this average event reporting and so on. So nowhere we say that no means no, no regulation. I think it's more just providing a more grounded uh approach that as grounded in science and evidence. And also again, like I said, and also actually in the end, we actually have um additional cultural action proposal is actually to have the community come together to discuss and propose what we call the candidates um air policies for we call the conditional response so that the community can discuss and what uh conditions, what policy would be appropriate. It's just in this way that we can have um you have multi hold open dialogue discussion in a low stake environment. So we hope we hope that this approach actually can lead to better consensus building process for better policy. Understood, Martin, can I take a quick, I really appreciate this question because I do think it gets to the heart of it. So um I think that our posture as a nation shouldn't change and it's changing. And I want to give you an analog which is like the internet. So I was in the national lab during basically the advent of the web and the internet, right? And we had this new stuff and like very quickly, like it was clear this was marginally different. We had computer worms, which we didn't have before. We had this notion of asymmetry, which means we are more vulnerable than countries that didn't use it. So clearly, it was marginally different, which we don't have for a I. So what do we do as a nation? We invested heavily in understanding it and building out capability. Like we became the leaders of the internet. We turn it into a discipline and like, and yes, of course, there's an entire cybersecurity. You know, like I've known Don for 20 years, he's like one of the preeminence researchers in cybersecurity. So it's not saying that we ignore it and ignore policy since we've become the leaders, we understand it the best and we make the most practical solutions that's not what's happening with A I I mean, the eu is doing whatever it's doing right. It's like, oh, there's this new thing we're regulate it and for us, for me, which has been so shocking is like academia wasn't even involved in the discussions. They've been silent. And so I'm saying yes, we need to regulate, yes, we need to understand, but we need to change our posture to engage research, invest understand the marginal differences. And then yes, let's do like we've done in the past, which I think is the right thing for the country to wrap this up and thank you for that to wrap this up. How are we feeling about where we go from here in in, in uh in one or two words, yours is cautiously optimistic. Yeah, Dawn Yes. So we need to focus on science and evidence and we need to advance our scientific understanding. Got it. All right. A round of applause, please. Thank you very much. I appreciate it. Thank you. Thank you. Thanks folks.

Liz Reid, Vice President, Search, Google Interviewer: Jeremy Kahn, Fortune

Brainstorm AI 2024: Google’s Strategies For Search

Liz. Thanks for being here with us. Thank you for having me. It's a great room. I can't see you at the back. But um I want to warn everyone, I'm gonna go to questions pretty early in the session. So please think of your questions. Um But first I want to ask you was, there's been a lot of reporting out there in the last few weeks that young people in particular are starting to Google things less. They're not going to traditional Google search. They're using some kind of A I chatbot or they're using these new A I enabled search engines, things like perplexity. What, how are you viewing this trend at Google? And how do you sort of want to trying to combat that potential loss of market share? Yeah. So I mean, there's always been different options with search and we welcome that. Um I think what we're really focused on with search is the user um and really helping that user find information. Um The truth is that there are not just billions of questions that people come to Google for. There's billions of questions people don't even ask and they don't ask because it's too hard, they don't ask because they're not confident they can get that answer. Um And so we really think about and search, how do we make it so that you can really ask anything you're curious about? And then it really feels effortless that it's so easy to do that. You don't think twice before you do that. And we think there's a huge opportunity ahead on that. Interesting. You mentioned, you know, people come to search because they need answers and they want quality information. And there's a lot of concern right now about what's happening to the information ecosystem on the web. There was already a problem sort of even before ja I exploded about search engine Optim A and whether that was degrading the quality of information that was available on the internet, then there's a, you know, a doubling of that concern with generative A I and the fear that's producing a lot of, you know, a I quote unquote swap that's out there. And how are you, how are you sort of seeking to address this problem at Google? Is it, is that our concern now that you're OK? You know, we have a great search tool, but it's got to be able to surface information that's, that's of value to people. So ensuring that we surface high quality information is, is definitely paramount and it's a huge reason that people have trusted search. Um This, this challenge has existed since the beginning of search, right? Um People have constantly created amazing content for the web and they've created poor quality um content in many ways. It's a, it's a cat and mouse game with those who wish to create um spam across. Um I would say A I generated content isn't by itself intrinsically bad, right? You can use A I um to really help people write and create and produce amazing information. Um But there's also this challenge of scale content abuse. Um And this is something we take very seriously. It's something we've continued to work on for many years. That's the level of spam and search is very low. Um And we'll stay focused on it because it's absolutely paramount. Um You've now rolled out these A I interviews, sorry A I overviews pretty widely that's created a lot of concern on publishers, the part of people and other businesses that really depend on search traffic coming from Google, you know, to their own websites where they seek to monetize that often through advertising. What, what are you saying to publishers about where this technology is heading? Because there's a lot of concern in the industry about, about what the effect of this is going to be? Yeah. So I think we have a long history in search and having a value exchange with the ecosystem and we've sent billions of clicks for a long time. Um But one of the things that there's sort of this common misconception is that people either want answers or they want links. Um That's not actually what we find from research. Um People wanna get started, but they actually really want to hear from other people. They're interested in their perspective, they're interested in their understanding of the space. And so we really tried to build A I overviews and continue to iterate them on a way that enables people to get started, but then continue to connect to the web. Um And we see that that's one of the things people love um about A I overviews is this ability to have that hybrid to understand and then to continue on in your journey. And I think that's only going to continue. We see that people, especially younger users really want to hear from the authentic voices, they want to hear from the people they trust and that's part of what makes it special, connecting to the web. And so that's at the heart of how we think about in the A I transformation still enabling that mission of connecting people to the best of the web. And how can actually A I make that even better in allowing people to connect to truly the best of the web? Are people still clicking links with A I overviews? We do see people click both within the A I overviews as well as on the rest of the page. And in fact, when we see A I overviews, we see that the clicks that go out to the web are of higher quality. People spend more time on the sites they go because these are really giving the information that they want. All right questions from the audience who's got a question right here. This gentleman has a question that hand went up very quickly. Um Wait till the mic gets to you if you can let us know who you are. Yeah, Josh Legasse with insignia. I didn't even know that was a thing. It just started popping up and I love using this. I'm glad to hear that. That's great to hear. Uh agree with the clicking back and forth. My question is, is it learning from me and feeding back answers to me or is that learning from everyone and feeding back answers to me or is there something else that's happening? Um So I think there's different levels of, of your question, there's sort of um what is the level of contextualization that we think about in these spaces? Um And there's some understanding of like, how do we interpret the query that can be more contextual. Um But in general, we use um throughout search, not just in A I reviews but broadly, um we try and learn from very privacy safe way um giving people controls, but we can learn a lot. This is how people learn these are the best sites to go to, right, is often based on user feedback. In addition to our high quality signals. And so we try to do it in a way that's very clear that gives people agency. Um but that we can actually improve based on the usage of the product. Other questions from the audience. Oh How about this woman back here? There we go, wait for the panel. Hi, I'm Julia Martinson with service. Now, um I mean, Google search has obviously been, you know, the best for a long time in terms of search and, and also providing the right results, you know, you know, ads kinda maybe, you know, for some more cumbersome, for more, some more enjoyable. Um I'm wondering with seeing how LL MS went also into more domain specific small language models. I wonder if um maybe, you know, a road map question, would it be ever, could we ever see a domain specific research basically where you're also kinda going down that level and, and making it more relevant for that specific domain you're in. Um I think we've continued to look at um uh different domains in different ways. I mean, if you come to search today, we have um a mode that enables you to do images, right? If you really want to go deep in images, um we have the ability to go deeper on shopping or on local or on others. Um And I think um that will continue, there will be cases where what really people want is a specialized um understanding of information, I also think with A I the ability to just um personalize the, the responses um evolves in ways that we don't traditionally think a lot of people previously with personalization with ranking, think of it as like, OK, I'm gonna re rank just the site. OK. Um But actually there's a lot of different ways um in which personalization can be useful. OK? Um it starts with basics like some people love video results and some people really do not. OK. Um Can you understand that? Um what's the level of understanding of what you want and responses? Are you an expert in that domain? And we should assume so or are you, you know, sort of trying to help your five year old with a concept that you don't actually know very much and explain it to her? And how can you do that? Um Are you on the middle of a long journey or on a shorter journey? Do you like to scan information quickly? Are you an audio learner? And then we should make it easier to understand information in that form? Um And so I think besides getting better within domains, there's just a huge opportunity to start to understand like what does it really mean to make information effortless if you go back to the mission? Um Most people like if they repeat the Google mission kind of start stop halfway through, they go organize the world's information and that's kind of where they stop. But there's this part at the end about universally accessible and useful. Like if we show you information that's organized, but it is not useful to you. It's not really that useful, right overall. And so I think there's an opportunity with A I to rethink what it means to give information in a way that's truly useful to you that truly feels effortless to consume and interact with. There's a lot of people really like the A I overviews and a lot of business is use Google search either on their websites to allow customers to search for things or they use it some internally on their own information. Is there the ability to kind of take the A I overview kind of technology and have that work on your own site or on your own information? Yeah. Um So there's a bunch of different offerings that Google Cloud has. Um uh And I will confess that I'm not the expert in Google Cloud. So you may want to get additional information on this. Um But at a high level, um we both have services that allow you to have your own data um basically ground a response in the data on your site. Um There's also a service that allows grounding with search so that you can get the best of sort of search his understanding of what's high quality information in your product. Um And so cloud offers a variety of different surfaces so that we can actually take some of the, the key underlying LL MS and understanding about how to get high quality information and make it available for others. I know there was at least one more audience question this gentleman right there. Let's get him in. Um, and then maybe I'll have time for one more of my own. But hi, my name is Matt. So I'm involved in running Ja I at mcdonald's, I'm curious, Google's famously brought search together in like one bar that you can do everything and, and yet now you're sort of splintering it again. You've got your Gemini product, you've got your Gemini app. Where do you see this going? Do you see it sort of coming back to one interface or do you see sort of the chat interface and the search interface? That's a great question. I swear. I didn't set you up for that because that was the other question I was going to ask is are the products going to blur together? You have Gemini as a separate product from search is that, are those going to be kind of recombined? I don't think we honestly know what the future has. We're pretty early in the journey. I do think that one of the things I think about a lot is how do you meet users where they are? And so, so if they come to search, can we give them the information the form they want if they choose to use Gemini. Can they do that? Um If you actually go back? And like long enough ago, people ask this question about youtube and Google's video search. They asked this question about Google Maps and um local search on search. Like, wait, where should I search for a restaurant on search or maps? Should we just merge search and maps together? Really? What you want to do is make it e like whatever is easiest for the user to come forth. Um The other thing that I would add in that space is that I think um you'd mentioned the keyword box, OK? Actually, search is evolving. So the keyword box is only a component, right? We introduce voice, we introduce hopefully all of you know what lens is. If not, you should come find me afterwards, that ability to search for what you see. And I think the space of information is going much more multimodality in its input. And the reason is because, you know, we're humans that existed a long, long time before technology and the way that we understand the world and we think about the world is based on a lot of different senses, right? It's based on what we see and what we hear and how we talk across. And so if you want to make information effortless, we're going to continue and search to think about what is the most natural way you would express if you had no constraints. If you assume tech was not a barrier. And you just asked your friend as easy as possible, how can you have that experience with search? So, so the tech really comes to you instead of you having to translate your need to the tech. Well, you've got this Google has, as everyone knows this department of justice case that's out there and there's just been this remedy phase and some of the suggested remedies involve the separation of search from other products. Are you guys actually sort of planning internally for, if you had to implement one of those remedies, what that would look like? So I don't have more really that I can comment on regards to this. What I will say about search is that we have consistently found that users choose search because they really find it the most helpful. And I think we're going to continue to really focus from, from my point of view on how do we help users? Got it. There's a lot of talk about sort of agentic A I and people may be using other agents increasingly, you might have people coming to Google search that are not actually human. It's, you know, you have an agent coming. How are you thinking about that? The idea that the user may not actually be a person, I think in general, you know, our, our goal at the end of the day is to help the user um as people interface or as as agents interface with search in different ways. Um I think we will always try and evolve it to the question about like, OK, how do you help the agent? So the agent helps the user. But at the end of the day, we're really focused on helping people. Um and we'll continue to innovate in that space, right? And can you give us any sense of, you know, road map is that we're going to see more multimodality? What are we going to see from search going forward? Um I think you will definitely see more multimodality. Um I think in general you will continue to see this evolution such that it feels like accessing information, feels just much easier, it feels more effortless, it feels like the question that you actually have in your mind. Um Is the question you get to ask. You don't have to break it down. You don't have to translate. You don't have to worry about what language you speak. You don't have to worry about how you learn best. Um And that search just becomes much more accessible in a new and exciting way. Great. Fantastic. That's all the time we have for uh for it today. But uh thank you to Liz and thank you all for listening.

Manohar Paluri, Vice President, Engineering for Gen AI, Meta Interviewer: Jason Del Rey, Fortune  

Brainstorm AI 2024: The Challenge of Defining AI Transparency

 Manohar Paluri, Vice President, Engineering for Gen AI, Meta Interviewer: Jason Del Rey, Fortune  

You're gonna help me with this pivot today after that performance. It's hard to match up, man. Yeah, we're gonna, we're gonna, we're gonna do our best here. So let, let's jump right in. Um Meta with Lama has a very distinct point of view on this topic. Can you explain why open your version of open? We won't get into the, is it really, really open? But um we could have those conversations outside but why your version of open and how do you think about the strength of open versus what some people see as the risks? It's a very often question that you get when you lead Lama and uh uh last year, when we faced this question, the answer wasn't super clear to us, but philosophically, uh we always believed open is the way to go all the way from open compute porch. And many of the things that meta has built uh Lama being the latest incarnation of it. I want to answer it in two ways. There is a particular way of thinking about A I development over the past four or five decades. Every time we overestimate what we can achieve with these models. And with the progress in A I, and we feel it's something to be enclosed and built upon and that's where the value is. But every time we hit a wall in unlocking General Intelligence, uh unlocking A I that actually need not be customized can be used out of the box. If you really look at the progress of A I today. In fact, even the 3.3 launch that we just did on Friday, there are multiple variations of those models uh and customizations of that that have already launched. And part of that reason is because it's open because people are able to build on top of it, customize it. So it's ridiculous for me to actually think of a world where we enclose A I and say you have an API for intelligence. Instead we, I genuinely believe the progress in A I over the coming decades is only going to happen if we actually go the open route where everybody can customize it, access it, um deploy various applications on top of it all the way from using an API but also customizing the weights. And that's where Lama uh vision stems from. And it actually has a tremendous amount of benefit for meta as well. So it has, it benefits the world, it benefits the developers, it benefits meta. So that's to me is the reason why we are open. So it's II I can understand seeing a A no brainer especially in the competitive landscape. And I and I get the, the speed to improvement and speed to innovation. Um We haven't really talked about the downsides and um you know, there, there are many out there that, you know, depending on your point of view or if you're a government official, you might see. Right. And so, you know, there's a report report in the fall that researchers linked uh to the Chinese military, we're building on top of an early version of L A. Um Again, I think the response was that's against our policy or our rules um still happening. So why should the world trusts meta with this open source approach with such powerful technology? When we know whether you know whatever your views of Chinese military are there, there could be actors out there who might not use it in the ways designed or intended. Um This particular one. so first and foremost, we should be very grounded on, on how we see the world uh with the data that we have and with the facts that we have, right, the competitive landscape of A I is pretty amazing in terms of the progress we have been able to make across the world. This is not a particular country. This is basically uh if you if you really look at the success of building the best models today, it has four components to it, this talent that you need to be able to build these models, there is compute which is flops that you put into these models, especially the flagship models are the largest large scale models that are actually used to distill and build very efficient models. And there is data, uh Lama at least has been built on publicly available data and license data. So unique data, the higher the quality, the better. And finally, the distribution which is aligning these models for various applications. If you have these four things, you are actually able to build the best technology and unlock value and the competitive nature is across uh the whole landscape across countries and continents. There is a huge uh competition in terms of building these best models. And to me taking a step back, if you want to play the long game of actually making sure this value accrues to everybody in the world. Again, you go back to how many people are building on top of it. How fast are you building on it and how responsibly and safely are you building it. And if you look at the operating systems that are available today, the safest operating systems are the open systems. So similarly, if you put out a model, there are thousands of people who are redeeming these models, there are thousands of customizations of these models that are much more efficient for hardware, much more safer for various applications. How is that going to happen if it's done by a single company or a single entity, it's going to only happen if everybody, all of the talent across the world are customizing this for applications and building a safer uh future and future that is more accessible to everyone. So that's, that's the real. So, so is the point that they, they're gonna always be use cases that we, we, we don't love but uh we're together, we're gonna build guard rails around that or am I, am I misinterpreting? I have a fundamental way of thinking about it, which is if you, if your router doesn't work, you actually don't go to the, you don't go to Linus Walters to actually complain about the router. You basically go to the manufacturer of the company to complain about the router. What is working. What is not similarly when I think about A I and these foundation models, that layer is not where you actually control or where you think about what is working or what is not what applications you actually enable on top of these models is where you have to have checks and balances. And to me that is the real place where you actually have significant value, both upside and downside. And we need to look at it from a responsibility and ethics point of view. That's why when you think about Lama, what we did is at a model level, it's so hard to compress all of the policies of the world into one model. It's not possible. So at a system level, when you build a system, you have this input and output guards, what does that enable you? That what enables you is that you can have the model but you have certain questions that you want to answer, you have certain answers that you want to curb. And that flexibility is what you need to give at the application level. And then the policies that are more regional cultural geopolitical can kick in and say, OK, this is how we are going to deploy these models and this is how we think about safety and responsibility. Let's shift gears real quick. I want to talk about some of the consumer applications of wa and meta A I so um meta A I started, you, you would know the exact timeline started showing up in all our apps, Facebook, Instagram, whatsapp, right in the search bar. Um What have you learned from? You know, I think what some people saw is just, you know, slap, slapping that on there and see seeing what happened like what, what is the feedback been? And then I wanna talk about the scale of, of that use case. So we've been blown away by the adoption of me. We are close to 600 million monthly actives at this point across all of the surfaces that you talked about. But at the heart of what we are, what we are trying to do is Lama is our engine that enables me, I I which is a universal assistant, but it also enables business A is wearables A I which is a Rayan meta glass that you can wear and you can invoke me I I, the assistant and with have a hands free intelligence experience uh created A I and many of the other A I that people can customize towards their liking, whether it's a particular fan following they have on a particular character and so on and so forth. So the goal for us with this foundation models is to enable all of these product experiences. So humans, human connection is stronger, human agent connection is stronger. And that's kind of like the at the heart of what we are trying to achieve building the future of human connection and the technology needed for it. And Lama becomes the centerpiece and med is our universal assistant experience. So it's been pretty amazing to see how med I gets used across all of this. How do you, I mean, because I, I've struggled to figure out what is the right U use case when I'm, you know, looking for, I don't know, an old high school friend or something. Um And that's probably not the right use case, but it's in that same search bar, right? And so um what, what are the use cases that have been most common or, or maybe surprising. So, so many of the uh so II I kind of have two or three that come to my mind because I always believe what you use is the best way to actually think about the product. Uh imagine is one of the experiences that you can use for in Medi. I, so you can actually imagine anything. And recently like three months ago, we had a birthday cake that I designed with my four year old daughter. You can imagine it. You can add a princess, you can add an Eiffel Tower, You can make it a strawberry or chocolate flavor. And then you have this custom cake and you're like, OK, this is pretty awesome. And then your building and then your cookie had. Yes. No, I didn't cook it. I had some, let's cook it. I wanted the daughter to be happy on birthday. Uh uh So there was one, you can actually invoke my I I in a group. Uh So if you think about how you uh uh use these models and interact with our assistants, you're all, you're, you're naturally in a conversation in a group and you can actually uh invoke me. I I to ask a particular latest score. So you're talking about a whatsapp group in whatsapp group. Uh So I'm an Indian cricket huge cricket fan. We're always talking about what's the latest score stuff? This is this group with, with uh my buddies and then we can ask me for the latest score. It will give you the latest information we can actually have uh some really fun roasting sessions uh for the groups on uh for the people on the group. So these are things that are now on me. You could open uh and you could definitely do it. Yeah, I see an editor who would gladly roast me right now. But these are some exact. So you want social plus this intelligence come together at the heart of it and you're already on whatsapp, you're already on Instagram, you're already on Facebook. So basically, we're bringing me I I and all of these A I in that place. So interactions with friends, interactions with businesses is natural and intelligent. Um I was curious about that 600 million number that I keep hearing about active users. So I asked meta A I like, can you like explain like what is an active user mean or what is an engaged user? And so I won't go through all the, we, we had a long back and forth kind of, you know, insightful. And at the end, uh meta, I told me that perhaps more than half of the 600 million might be not explicit uh engagements where I know I'm engaging A I, but maybe a meta A I is working behind the scenes on a result uh on a search result. And um I was just curious is, is that accurate? And um because I, I know, you know, all the companies are very proud of these active user numbers. And so Um Yeah, I'm, I'm curious, I'm curious for your take on what, whether, whether that sounds about right? That maybe half of that number is sort of not explicit uses, but A I working behind the scenes, let me answer that. Uh So the two answers to it, one is uh technology when it works should be seamless. So when you are actually doing a search query or when you're actually asking a particular answer, you're actually trying to do a thing, whether it's on our app, any of the surfaces, whether it's whatsapp Facebook, you're not necessarily trying to say I want to use search, I want to use Medii, I want to use this. You're trying to get a particular job done. You're trying to solve a particular problem from that context. We're trying to make sure Medi I can understand what your query is and if it requires a language model or if it requires a real time query or if it requires a frame graph access, we are able to do that seamlessly for you. So you can actually get the job done in a much better way than before. That's how we think about it. And going back to the question that you are asking, uh it's really hard to break down which particular intent was search only which particular intent was accessing a frame graph or looking at a particular subs string or asking the language model for latest information and so on. So all of this together is what we are trying to unlock with me. I was going to say when you said two different parts, I thought, I thought that was going to be a really wrong non answer. But we're really saying we should just trust meta A I's answer. Maybe we're, we're, we're actually, the technology is definitely improving day by day. And that's one of the other things if you look at the launches, we have done over the last year, we basically did five launches all the way from Lama three family to 3.3. And if you look at where we started the beginning of the year and towards the end of the year, we have more than 600 million variants of a Lama that people have built on the 600 million monthly actives that we talked about. And me, I was in the fledgling uh part, we just started thinking about me. I, I, so going from uh product experience that doesn't exist to a product experience that's used by these many people. It's pretty amazing. I think we're all at the time. Thank you so much, Mano. Thank you. Thank you for having me.

Leigh Marie Braswell, Partner, Kleiner Perkins, Sarah Guo, Founder and Managing Partner, Conviction, Rebecca Kaden, Managing Partner, Union Square Ventures, Allie Garfinkle, Fortune

Brainstorm AI 2024: Beyond Hype - Seeking AI’s Enduring Value

Andrew Feldman, Co-founder and CEO, Cerebras Systems, Mark Papermaster, Executive Vice President and Chief Technology Officer, Advanced Micro Devices (AMD), Sandra Rivera, Chief Executive Officer, Altera, Sharon Goldman, Fortune

Brainstorm AI 2024: Chipping Away at the Competition 

Andrew Feldman, Cerebras Systems, Mark Papermaster, Advanced Micro Devices (AMD), Sandra Rivera, Altera, Sharon Goldman, Fortune

Jessica Powell, Co-founder and CEO, AudioShake 

Brainstorm AI 2024: A New Tool to Master The Mix

Hi, thanks all for coming. Uh I'm Jessica Powell, the CEO and co-founder of Audios Shake and I am waiting for the slides to start. Oh, there we go. All right. Um So today I'm gonna talk to you about sound, but before we do that, we're gonna do a sound experiment. So, in just a moment, I'm gonna ask you all to clap and to cheer as if you were at a sports stadium or a concert and I'm gonna talk and can we get some music? Great. Ok. So louder. Can you hear me? Can you? Great. I really just wanted to applause. Um Great. So now let's see if I had asked you what the person 2 ft from, you had said during that time, you would have no idea or if it had been a concert and you would actually wanted to hear what was happening. Like maybe more you wanted to hear the music and the crowd, you'd be stuck and that's generally how we experience sound in the real world, right? Think about being in a noisy bar or being uh you're on the street and you're recording a video and all of a sudden a, an ambulance comes by or a police siren or since we're in Silicon Valley, let's say you're developing an application that requires voice input. Well, if your user is in a call center on a noisy street corner or simply as toddlers running around in the background, good luck getting the input that you need. But what if that didn't have to be the case? What if we could actually extract what we needed to hear? So let's listen to something that's very, very noisy like this. And now let's isolate the voice and lift off of the space of returning to the playstation. So this is what we do at audios shake. We make sound work better for everyone. We split sound into its different components in order to make audio editable accessible and useful. We're a core audio infrastructure company a bit like a Dolby. Now, this is actually pretty hard to do and let me show you why this is an image of an audio recording and you probably don't know. Is this a whole bunch of people speaking in a crowded room? Is this a music recording? Is it a bunch of sound effects in a movie? And all I'm gonna tell you is that the Y axis is the frequency and the X axis is time and that's about all you're gonna get. Now, what I'm gonna do is I'm actually gonna color this in for you. So you can see what's actually going on. So this is a song recording. Um And you can see we've got some vocals, some drums, some bass collection of other instruments. And again, this is not actually how sound looks. You have no way of actually knowing what's in these different parts. Now, recording engineers have a hack for this developed back in the sixties around the time of the Beatles, which is multi track recording. You send the, uh you know, the vocalist into the studio to lay down their track, then the drummer, then the bass player and you now have all these granular elements that you can do granular audio editing with. You can remix the track, you could put those different sound objects in different perceptual fields. The bass here, the drum here to make a surround sound mix, you could correct imbalances in the audio all because you have those parts. But real World sound doesn't come to us like this, right? It's, it's more like what we just experienced in our little experiment. And so we have to turn to A I and deep learning to help us fix this problem and be able to get at these individual parts. So let me show you how we do that. And in fact, before I move slides, if you look here, uh if you look at the yellow and the orange, the drums and the bass, this is a really great example. So you can see that these are not in one tidy band, they're all over the place, right? They're overlapping with all the other frequencies as well. And remember, we don't see any color in this image. So all we can see we have is essentially sound patterns. So what we do is we train on hundreds of thousands of these parts which are called stems and let's say drums in order to learn what these sound patterns are. So let's say we can, then we can then take an audio recording that we've never seen before and recognize that there are drums in there. This, this matches to the sound pattern of a drum. So we figured out what something is, but now we need to disentangle it from everything else. So let's say we want to do that with the base. We want to isolate the base from the recording, think of an image editor. So what we're gonna do is we're gonna leave everything that we've all the pixels that we've identified as corresponding to the base. We're gonna leave those there and everything that's not the base. We're gonna black out and essentially, we're gonna silence that audio and that's gonna leave us with the base. Now, this might sound, as my parents said to me like a very weird hobby. Um And, and, but actually, it's very, very practical and audio tech works across a ton of entertainment and speech work flows today. So in film and TV, we can isolate dialogue, music and effects to take old film and open it up to localization and other languages or remove on set uh wind and noise in music. We can separate the different instrument stems for remixing or immersive sound in sports. We can boost what a player is saying on the field or remove music from the environment so that teams and leagues can avoid millions of dollars in copyright signs in transcription and captioning. Our tech is used to isolate the voice before it goes through transcription so that you get much higher accuracy. And in generative A I were used by some of the world's largest foundational models to learn human conversation. So for example, disentangling overlapping speakers or extracting the uh huhs and ya that make human conversation human. So let me show you a quick demo of what this all looks like in practice and the s what there a lot of dream come true. Innit. It really did. Did you uh did you, we talked about love and coming into a game like this? It's the magnitude of this game. Massive game. Did, was there a a hangover at all? The result? And the that coming to our lane came to earth, you couldn't be more mistaken but here to help you by deactivating them. The Lord. Yeah, don't have to trust an goodness said Noah, I need at an airport. So I think I came to the airport like 30 minutes before boarding maximum. I wanna walk right back here. Enjoy the lounge. That's why I call the lounge. In fact, there's going to be so much more that you're going to be able to power with sound separation in the future and in the near future. So we've already heard a little bit today about music in music. You're going to be able to remix and reimagine everything and all content including including c is going to be able to be made immersive both through sound separation and things that are happening on the, on the vision side as well. Uh We're also going to be able to do all kinds of measurement and analysis thanks to sound separation. So imagine being able to measure the degradation or the health of an ecological environment around the presence or the absence and the change over time in different kinds of animal sounds and in generative A I, we're going to be able to the foundation models in fact, are going to go deeper and deeper into audio and to the point where they're going to be able to understand audio, the way we do. So all the complexity, the way any human here could have said that was crowd noise and there was a dog barking over there. Computers will be able to do that too and they're going to be able to learn to talk like us as well, which is going to change so many workflows from the factory floor through to kiosks in the amusement park. Finally, because of advances, not just in A I but also chips and hardware, we're going to be able to solve the noisy bar problem. Now, one big technical challenge in all this is speed. So in sound separation, we are taking large audio files, running them through large deep learning models and then output large high resolution audio files. And that's why for the past year, audio shake has been working on building faster and faster models. And today we're launching here at Fortune uh the audio shake voice SDK, which will isolate voice in real time for noisy backgrounds, uh streaming capable on edge devices. And I think while we were all talking and doing our experiment in the back, they were running exactly that or we're about to find out if they did. Uh Can you play the full mix of what that room sounded like when everyone was cheering? All right. And what did you isolate louder? Can you hear me? Can you hear me? Right? I should have done something like you all want a free car or something like that. Um So if you go out today into the noisy hall or in a street corner later or at a bar tonight, watching a DJ remix of track, know that all these experiences are going to be enhanced by sound separation and audio shake is gonna help make it possible. Thank you.

May Habib, Co-founder and CEO, Writer Sujay Jaswa, Co-founder and Managing Partner, WndrCo Jeffrey Katzenberg, Founding Partner, WndrCo, Andrew Nusca, Fortune

Brainstorm AI 2024: The Business of Better Storytelling

Hey, folks. Thanks for being here with us. OK. May I want to start with you? I saw that your company is all about business writing. I'm a business writer. So uh help gonna replace me. Well, that's, that's where we got our start. Um We are the only generative A I company that is productive on top of every step, function change in the capabilities of the, the transformer. We were founded four years ago. Um And that's what models were good at then. Now LL MS and natural language really are the, the gateway to intelligence. And now we say if you can write it, if you can describe it, you can build it. And so we have pioneered the enterprise generative A I category by combining LMS with retrieval and guardrails and orchestration. And it's that full end to end solution that enterprises need. And you know, that's what we've built over the last few years. So I'm safe is what you're telling me. I don't know. I'm right here, mate. All right. Um Su J I want to come to you next. You've invested in this company. Um Tell us a little, remind us what wonder coasts thesis is and how write it fits in. Yeah, of course. All right. So, first off, we were incredibly lucky to invest in May. Uh She's crushed it since we invested. She raised it like four X devaluation a year later. I mean, it's kind of wild what she's done. I'll talk about that in a second. Um What wonder co does is we create a company every year, about one company a year. Uh We make about 5 to 7 venture investments a year and we make 15 seed investments every year. The focus of the firm from the beginning. This is kind of like an unknown thing we were talking about it earlier is from the beginning two thirds of the money that we had went into software at that point in time, one third went into digital media in 2020. We said, you know what, let's just move entirely to software. And so since 2020 we've been 100% software, but we've been doing software since the, since the get go. Yeah, for sure, Jeffrey, you have your pick of what you want to invest in. Why A I stuff right now, what are the opportunities and risks in going in that direction for you? Well, I mean, I think that uh again, as CJ says, our focus is really um uh been productivity tools. How do you uh take IA I and create applications that make people more effective, more efficient, more productive, more creative at work. Um, and, um, the, there's certainly the, you know, the, the big guys that are out there who are, uh, I think doing a pretty world changing, uh, things in terms of, you know, what, what's the future look like and sort of the larger future of A I, but what we've tried to do is just sort of be boots on the ground on a very practical basis. And you have companies like writer today who can instantly deliver productivity and real ro i to companies of things that are essential to their day to day business and hers around storytelling, which is in particular why I think we were excited and why we related so much to her to writer and to her mission from the outset. Yeah. Yeah. May I want to press you a little bit on this? May it may be, I'm a replaced or not thing. We've been talking a great deal over this event as to the role of humanity and, and creativity vis a vis the tools. So how do you look at that? Especially given what you do, where do we fit in as humans? We're tools. Also humans are tools and I know that sounds nuts. But we are moving to a world of autonomous action for A I where we're going to be nudged when we're needed. Uh And uh across the street, uh we've got our customer conference going on where Uber sales force. Victoria Secret are, are walking folks through what is truly possible with, with a Gente A I using our, our tooling. But the guardrails are so important and you know, folks are so obsessed with, with LL MS. And that's one piece. The part that folks really miss is and, and this is part of our storytelling, this is now very meta, but this is really what's happening in the enterprise, you know, to make A I work. You absolutely have to get your business process, your business logic, your data in order, right? And marry that with the LL MS whose reasoning capabilities really are increasing. We build our own LL MS. We are pushing the envelope on the architecture of the transformer to enable the kinds of complex orchestration of, of work that truly is possible. But it is no question so confusing for the enterprise right now. And it's just mass delusion that you can get to transformational generative A I working just with the hyper scalars Jeffrey. I want to come back to you. Same question essentially around where the humans fit in. I think about DreamWorks. You started it in 94 and would you start it the same way now? Given the preponderance of all these tools? No, but I've, I've lived through these transformations before. Um I'm I'm I was born in the Jurassic era and we used to make animated movies by hand drawn cell animation and um and went through that transformation into um uh the resource and assets of technology as the assets, two artists, two animators uh and then obviously into what was the most transformative part, which came in the sort of mid nineties, which was CG I animation, which obviously has become, I don't know, four out of the top 10 movies this year. So I've seen this story before it gets retold time and again, and each set, each time there are these new sets of tools, it's a new paint brush, it's a new, uh does anybody know what a typewriter is? I've heard of it. Um And so I actually think these are phenomenal tools. I think they are gonna be extraordinarily empowering. I said, I don't know. Now it's a year ago where, uh you know, when I, you know, left the business now, nine years ago, it took 500 artists, five years and $200 million to make a movie. And, you know, that is already on this extraordinary curve down. It doesn't mean it's less creative and it doesn't mean that it doesn't require a great storyteller and a great vision. It does. It's just we're giving a new set of tools to them to actually realize um and make these dreams come true. Right. Right. You know what, 11 thing to think about just to, just to contextualize it because I think this feels scary because of how um most, most of the people that we all interact with, have the application of creativity as part of their job. But if you look back at like, if you just go back, like to the time of the American revolution, the industry that 75% of American workers worked in was agriculture. And you look at, you look today and it's like 2 to 3%. And it's not because we're producing or consuming less food per capita, in fact, probably the opposite, right. It's, it's because tools allowed us to kind of supercharge things and humans could go away from work that they would, you know, kind of the parts of the work that were drudgery and move to things where we could get to higher and higher orders. And I think likely it's going to be the same thing with A I tools. So do you think that, I mean, you're looking at the whole landscape and trying to choose, make your bets? Right. Um Do you think that we're going to lose entire industries or entire companies because of these dramatic changes we've, we've articulated at the individual level now in this conversation, how do you look at it at a bigger scale? Look, the big debate is whether A G I is gonna happen in the near term or whether it's gonna take a long time and you know, there are people smarter than me that would tell you that A G is around the corner, I think where we're betting as wonder co as we think we're going to see, you know, scaling limits to the development of the current generation of models and algorithms. And we think A G is a long way away. And if we're right, then I think what you're going to see is industries are just going to get better because they have data, they can make things better, drudgery can go down. But all of us are going to be fine if we're wrong and A G I is around the corner, all bets are off, got it. I'm gonna come to the audience for a question in just a minute, but I have one more for you, Jeffrey. Um You spent a great deal of the last year being the co chair of the Biden campaign. I would love to know what you learned about how Silicon Valley and technologists are thinking about A I in that context. What did you learn campaigning for, for President Biden? Not that much next, fair enough to the audience. They won, they won, we lost their bat their ball and I'm, I'm rooting for them to play. Great. Seriously, like I, I, you know, this is the way I've been doing this for 50 years. And, you know, I've been been on the winning team and I've been on the losing team. I should be on the winning team, but it is their turn at bat. And honestly, for the future of this country and for the things that I think means so much to, you know, to many of us II, I am rooting for their success in every single respect. I, I hope the changes that they bring make us a better, stronger, more successful country. Yeah. Right on. All right. A question from the audience. Yes. Right over here. Hi, Steven Steven Wolf Berea from Aon. Co May. Huge congrats love everything you've been doing with. Writer. Your issue about data governance is so important and so many of these companies are just not A I, right? Because of the data issue, where do you see some of the challenges? Is it more on the CTO technical side where they're kind of um not made here and they're kind of being protective or is it on the business side where maybe it's the CMO but they really don't understand everything that you're talking about. Yeah, so maybe 22 challenges both on the business side and and on the tech side. So, you know, on on the business side, um it is AAA real interesting um challenge. If you think about the industrial revolution, we all those technologies existed, but we didn't get that big, big change in productivity until we got the assembly line. And so much of what is required is this assembly, this guard railing of the work. Give you an example. Large fortune 10 asset manager. How do you do investor reporting? Right? You've got five mil on a credit fund and you've got 10 mil on a private equity fund. Literally, how do we do reporting? Nobody knows, right. Someone outsources it to a si for the tune of 200 mil a year. But there's no process and to get generative A I against that process, the, the data, right needs to follow the, the, the business logic. And so the human effort to actually go in and diagram out all of this stuff. That's data too, right? The examples of how we've done it before, how we don't like to do it, how we want to do it, that's data and how you make that into continuous pipelines. That's a technology problem right now. And so, you know, folks paint a very wide brush of data is challenging for generative A I but like let's get specific and it's all very use case driven. Yeah, that's well, well said, thank you may. Well, let's go for one more, please. Yes, way over here. Hey there, good to see you. Hi, Laura Yeller from playground. Uh A number of I know you've been talking about enterprise productivity and I think we agree that is wonderful with the appropriate guardrails. But a number of artists and writers who are writing nonfiction and more creative works have had successful pieces and then have been approached by A I companies to write the Sequels and use A I to generate those based on the characters and the initial work. And they have been told by other friends who are creative that that would be terrible. It would dilute their brand and using A I for that would make it look like they didn't have that creativity going forward. What do you advise folks who are being given offers to develop a I cogenerated content? I'm, I'm actually not uh I'm not sure. So I don't wanna. But what I will say is, is that I, I have uh many uh friends of mine are uh among the leading writers, creators and showrunners in Hollywood today. Um And I would say almost across the board, they have all talked about how uh the A I tools today have been helpful to them. They have seen them as a resource and as uh an asset, they have made them uh uh more productive. They've been able to, to uh widen uh uh the uh diversity of their work, the quality of their work. So, again, I'm just, I'm only telling you what I hear from, from I would say among the most talented showrunner creators today, how they find these tools, um an amazing resource a for them and it's not constraining them. It's inspiring them. Can I build on that, Jeffrey? Go for it. We've got, I'm sorry to interrupt you. Wrap it. The answer for Jeffrey is no. Forgive me. May please a round of applause. Thank you for the OK.

Ed Newton-Rex, Chief Executive Officer, Fairly Trained Carol E. Reiley, Co-founder and CEO, DeepMusic.ai Moderator: Verne Kopytoff, Fortune

Brainstorm AI 2024: AI Music Finds Its Groove 

And a big thank you to our guests. We're here to talk about music. And before we ask, uh you know, what your organizations do, I wanted to set the stage and ask um uh A I can, can A I music be groundbreaking? Can it be as uh unique as the Beatles or Jimi Hendrix or Beethoven? It's a big start. I start with you Carol. Start up with a Yeah, an easy question. You know, I think um when I think about A I, it was um like through the full entire stack because I think the obvious one is like music generation and helping artists compose and write songs. Um Then I started like digging into like music production crowds and you looked at like the labels and then like the, the front of house marketing tools. And I, you know, I think the A I is, we agree is like the uh the superpower. So how do you use A I to become super creative? So for me, I was really looking at how do artists become super creative because I view them as like the most creative humans on earth who have made this their profession, how would they do it? Um And then see how that impacts um society and humanity later on. Um I think it has great potential. And I think for the last few years, I've been digging into exactly that question on where is the value and how can we make the pie bigger? Um And, you know, recent surveys have also showed that uh the number one job that young people want to do is to be an influencer. So how does that type of creativity also utilize A I to create content to put that out? What about you? What's your perspective? I, I think I've seen, I've seen innovative things be done with A I Music. Absolutely. When I was at Stability A I, our audio team released a product called Stable Audio and it was basically a sample generator and, you know, you could get it to do some pretty wacky things. Like one of my favorite things that one of our team produced was I think the prompt he used was hammering wood at 128 beats per minute. And like it came out with this weird combination of like a kind of hammering sound effect, like, but in like a drum beat form and I think sample generation in general is something where, where this tech is already being adopted and you can get some innovative stuff out of it. I do, you know, I think unfortunately though, like a lot of what is produced with a lot of the systems out there at the moment is incredibly derivative, right? People call it slop and I think that's like a pretty good term for it. So I think that the the majority is not going to be, you know, musically innovative in that way. I think uh well, I just also follow up and just also push that like A I could, the same thing can be said about humans. I think a lot of the stuff that we also produce at least to start with is also incredibly derivative or trying to copy cord. So I'm, I agree. I think that maybe A I could be used as a challenger to, to push us to be better. Um could be one lens to also look at. But I do feel like it's an inflection point where it could be a lot of sloth and, and derivative music and A I might just streamline everything to be the same derivative music. So, so I'm looking for us to really push that boundary. Well, tell us what your organization does. Uh Deep music. A I? Um So, uh you know, I think uh five years ago, II, I, uh I just sold my company which was in self driving cars which had a large impact on society and I was looking for what could be the next big disruption and I went through my own existential crisis. I'm like, what if a I reached the point where um it could replace human creature. These because we sat on panels and I always had said, if your job is in the creative, you're gonna be safe. And I was like, what happens if that assumption isn't true? Like, what would that look like? And what if we were to let a I replace the creative side, the stuff that I enjoy doing? What would that look like or how would we, how could I find a way to like, utilize and lean into a I so that um that could become a new area or like, so, so I really wanted to partner with artists and um people who were more creative than me. I'm an entrepreneur and I think that's like one set of creativity, but I couldn't figure out how to like systematically break down how I went through my creative process. I know that sometimes I just get inspired. The light went on my head, but I started talking to different types of creatives trying to figure out how they went through their process and had um landed with the San Francisco Symphony. I was working with them and composers and musicians trying to understand that. And so I created this nonprofit with one of my dear friends from grad school who is a artist, um Hilary Hahn, like it was during COVID, so she was not touring. And so we had like a three hour beautiful conversation about what happens if A I was to disrupt her career. Um because I had seen that in the self driving car space where eight graduate students from Stanford, we were able to put a car out in three months um on the streets driving around mountain view. So I would say I could see the acceleration of A I and I knew that it was coming for creative soon and wanted to help the artists not be reactionary. But if they were to lean into it and band together and understand the tools like and be part of the companies that were building these tools, how would, how would that change like the course of humanity? That was my 2018 revelation. And what I was been trying to unpack since then. Fantastic and tell us about Philly trying uh 30 well, 30 trains a nonprofit I set up about a year ago. Um I worked in, I've worked in Generative A I since 2010. Uh And for, for like a day for about 12 years uh since I was joined Generative A I to about 2022 people didn't train their models on copyrighted work without a license. Everyone knew it was illegal. No one did it. They, they trained research models but they would never commercialize it. But then in 2022 a few big companies, you've all heard of started releasing commercial models and threw caution to the wind and I ended up working at another generative A I company and, and eventually I quit that company because, because I, I thoroughly object to this unlicensed training. I think it's illegal and I think it's unethical. And so I set up fairly trained to, you know, we're a non profit. What we do is we certify generative A I companies that don't take that approach, we certify generative A I companies that don't claim that they can scrape the life's work of the world's creators and train vast commercial models on it. The good guys. Well, I think so, but everyone thinks they're the good guys, you know, and then you've also uh supported or uh uh had your petition signed by 12,000 creatives including this, uh the Cures, Robert Smith and radiohead singer Tom York, uh calling for uh compensation for artists. Yeah, it's tripled. We have 36,000. Um Yeah, this was a statement we put out like a month ago and it's a 29 word statement. The wording, it's still available for signatures. If you happen to agree with this, we're in San Francisco. So you probably don't. But if you happen to agree with this, the statement is the unlicensed use of creative works for generative A I training is a major unjust threat to the, to the livelihoods of the people behind those works and must not be permitted. That's the entire thing. It's been signed by a lot of, a lot of creators including you know, a lot of well known creators, but more importantly, like thousands upon thousands of working creative professionals who you probably haven't heard of and whose lives are being impacted by this uh this right now. So if you do agree, please go and sign it. It's still open, it's still rising. And now the music industry has sued a couple A I uh music engines. Um What should we think about that? Carol? Um IIII I agree that they need to be compensated. Acknowledged. I, I do think that uh technology has disrupted the music industry many times over from like sampling electronic music. Uh And this is another new tool that is coming and they need to be, I think, fairly compensated. Um I think it's great that they're standing up. Um And, and doing something about it. Um Another approach is II I think it's also interesting that some artists have openly model, like create models of their voice so that the public can utilize it and they can rev share things. So I, I feel like there's new business models that need to be created. We need to be experimenting with them, but I do 100% agree that the rev needs to be split. Um They, they need to feel fairly, not just monetarily compensated, but I think acknowledged for their work. Um There needs to be ways to like spot if that has been um not upheld. I think what Ed is doing is fantastic. Um And I, I think really banding the artist together to have like a community and a place to talk and have conferences like this is important. And what, what kind of uh compensation scheme do you envision? I? Is there a right one? I don't think there's a right one. Actually, I think there's like, I think often A I companies will turn around and say, well, come on then rights holders creators like, you know, what's the silver bullet for how this should work? And I don't think there has to be one like the beauty of licensing and licensing is clearly the solution I think here, right? The beauty of licensing is that like two parties can come together and figure out a deal and figure out what works for them. That's not going to be the same for everyone. Someone, one person is going to want an upfront payment, one person is going to want a revenue share stability A I when I, so I did licensing deals like this with creators and you know, and, and you know, we, we, we put in place a revenue share with, with music creators whose music we trained on. Um you know, and I think I think that kind of model can be great because no one really knows how much money a lot of these things are going to make, but a revenue share can work really well. Um There, there is this like there is this argument that A I companies often give in legal filings as well, you know, where they will say, well, you know, it, it's kind of, it's not feasible that we could pay everyone. These payments would be so small, like, kind of what's the point of paying people and to that, I say you've obviously not been a creator yourself and like, got small royalty checks from Spotify or whatever it is, right? Like, you know, like just because the payment is small doesn't mean a creator doesn't want it and it doesn't give you the right to take their work without permission. So that, that's my view. I think you've got to start from licensing and consent and you work from there and if people don't want to be involved because you're not offering them enough money, then you don't use their work. And it's as simple as that. Uh Carol, what will a I ultimately mean for the jobs of, of, of musicians, do you think? Got it? I do think, I think a lot about like the future of work and where people go and how we utilize our time. Um You know, it's, it's undeniable that it will be disrupted. There'd be a lot of jobs replaced, particularly um things related to maybe like music for commercials or royalty free music. Um That, that type of thing. I, I don't know how much joy those artists have created music, but there is a paycheck there. And, and I, I do think that, like, um those lesser creative, like, uh those type of jobs might be uh more easily replaced because it will be cheaper using like a I generated music or you could have personalized music in the own your own lyrics. Um But I do think there's always room for like true creators or like, I, I do think that more people will be jumping in. I feel like uh music seems, I, I've worked in many different applications and music seems to be really one that resonates with the majority of the public. I feel like music touches people. It is um gets them through hard times or they themselves want to express themselves a certain way. So I think there'll be more artists than ever before. Um And people would be able to use the tools and I, I think what we need to work on now is really the interface of these tools. Uh You know, if you guys have played around with a lot of the, the, the tools that are available, I think they're fun for a while, but the stickiness level doesn't seem to quite be there. And what I'm interested in is, you know, how do we get people to come back to the tools? Um You know, I think there is something with personalization. People love like A I bias for things that are personal for their life or their, their album um photo album and they want personalized music, but it doesn't seem to be something that has any stickiness right now. So I think a lot of work needs to be done on the interface side from and ed, what percent of the music industry do you think is doing it right now? Following copyright that you believe they should? Well, I mean, in the A I world there are lots of companies who are licensing all their training data. So fairly trained have certified 16 music A I music companies, companies involved in A I music who have, who don't use copyright work without a license. It's the, it's the absolute minority of companies in the music space who are training on the world's music without permission. And, you know, the two most prominent Asuna and Yio and they've both been sued for huge amounts of money by the record labels and I'm, I'm sure if other people follow their footsteps, they'll be sued as well. Um, so, yeah, a lot of A I music companies are doing the right thing. Like I encourage you to check them out. They're, you know, they're on our fairly trained website. Um, you know, II, I agree with Carol. I think that there are good use cases. Um, for some A I music, you know, I think, you know, one thing we saw recently, you know, was like Randy Travis, the country singer using, who had a stroke using, uh, using A I voice models to recreate his voice and be able to release new music. And there are these sort of empowering stories. But unfortunately, I think right now it's pretty unavoidable that the main use case of A I music will be essentially the commodification of music. You know, there isn't text to music models. I agree with Carol. I mean, that, that, that retention, that that kind of that real use case for a long time, people were thinking, where's the mid journey for music? Right? Loads of people use that and then a couple came along, you know, and how many songs about your trip to the coffee shop? Do you really want honestly? Like text and music is not, is not the answer. Um You know, so I think, I think there's a lot of work to do in the A I music world to find the really, the really good application. And unfortunately, as I say, I think commodification is gonna, gonna be the most scalable one and that's something I'm not, I'm not thrilled about personally. So, well, let's look ahead five years from now. Um Do you think uh we'll have an A I Taylor Swift? Is that possible Carol? Um I mean, I don't, I don't, I would say that the goal is never to like mimic the artist, right? I think Taylor Swift is a powerhouse and her job is which I get no store on her caliber on her love. Yeah, I, I think what is so amazing about her is that she is the entire full stack. It seems like from um songwriter to like CEO to like paying those bonuses and um to all the, the crew and everyone in running this company. Um So she's like an empire and I think that is interesting to see how she markets herself, how she connects with the community. Um And I think a I can help maybe an artist that uh doesn't have her efficiency, productivity, like really like use her as a bismol to learn all the different aspects of her job and maybe utilize A I if you don't have as powerful as a team um to be able to do some of that. Um So while that, that is interesting um to have we, as we've seen like a I avatars, um you can go to hologram concerts. Um You know, they like, I think Kim KK, Kim Kardashian for instance is interesting because she puts out so much content. So can you imagine like a superstar like Marilyn Monroe being um like just the the team behind her constantly putting out 24 7 content with her seat, see things and doing deep fakes. And now all the eyeballs of our world are watching like only five characters. So I see. So I see that future happening where there's only a few a few superstars and it becomes harder for independent artists to break into that level. And get the, to get the average consumer's attention. Um And I think what the music is working on uh besides the the machine of oops of the, the artist is what are new ways to utilize A I that you can't do by yourself. So we are looking at like neuroscience um um and brain activity, live performances and um and, and getting the mood of a crowd to affect the performance uh during the stage live composition, um visuals for, for music, robots, dancing with ballerinas, you know, I feel like new types of art forms and trying to figure out uh what is interesting that is in a gimmick that has fame power, that could be um a new way to appreciate art um as well as behind the scenes for like dynamic pricing and helping, helping um some of these like art houses thrive in this, this phase. Great. I want to thank our panelists for joining us and talking about the future of music. I'm looking forward to how it turns out. I hope the musicians get their due. Thank you very much.

Lan Guan, Chief AI Officer and Head, Center for Advanced AI, Accenture, Mahesh Saptharishi , Executive Vice President and Chief Technology Officer, Motorola Solutions, Jeremy Kahn , Fortune

Brainstorm AI 2024: The Art of Scaling AI

Presenter: Amos Miller, Founder and CEO, Glidance

Brainstorm AI 2024: Now You’re Gliding - Independence Through AI

Hi, everyone. Are you here? Great. Uh Thank you. I'm Amos Miller. I'm here with Paul Willard from Grep VC, one of our early investors. And before we start, I wanted to ask if you would close your eyes for a moment. Don't be afraid. It's ok. And when I snap my fingers, you can open them again. But what if I never snap? You've had sight for most of your life. You saw the smile of a loved one. You saw the colors of spring and now you feel the chill of uncertainty. Will you ever again be able to take that walk on an autumn day? Will you ever feel independent again? I snapped. You can open. So I'm Amos Miller. I'm the founder and CEO of guidance. I'm also a father, a friend. I'm also blind. Uh when I lost my sight in my early twenties, I experienced that chill. I had to go and learn how to navigate and get around independently as a blind person while uh on campus, learning computer science, going through the halls of London Business School, walking my daughter to kindergarten, traveling the world with Microsoft. It was stressful. It was hard and all that time. I am working at Microsoft really building one of some of the most advanced technologies on the planet. NASA is landing rovers on Mars. And I knew that I had to look into how do we tap into this advancements, these advancements in technology to really revolutionize independent mobility for blind people. My work led me to an investment by Satya Adela, the CEO of Microsoft uh where we developed one of the leading navigation technologies for the blind called Soundscape. It's an amazing innovation, but still we are not getting more blind people out. And about how many blind people did you see today? Statistically one in 100 people, you should see more blind people out and about. So I really started to real look at this problem and I realized I wanna describe to you what the problem is here over the in the US, in the in the world. There are 43 million people who are blind, the great majority lose their sight later in life. For macular degeneration, glaucoma diabetes. The numbers are expected to grow by 55% in the next 30 years. And here's the issue. It doesn't really matter how much technology we throw at them. As long as they're using a white cane or a guide dog, it's not going to work. There are only 2% of blind people who actually use canes and dogs. And why is that resources, money time. It takes five years and amount almost $100,000 to train one blind person to use a white cane or a guide dog and have them confidently get around the LA in the U si. Think the the presenter uh told there's 1 million blind people. Actually, if you count low vision, 2,000,003 million, the the largest guide dog school trained 225 dogs last year, it's completely out of whack. The system is simply unable to cope with the volume and is failing millions of people who want to get out. And about. So imagine if we had a new kind of mobility aid, a mobility aid that was quick to learn, easy to use and affordable an intelligent guide that was with you keeping you safe guiding you around, you will be back on your feet in no time. Our work at glance is pioneering the era of intelligent guides glide. As you can see on the screen is an intelligent guide with two wheels on the ground. A long handle, you hold on to the handle, you nudge the device forward. It uses autonomous driving technology to guide the person. Keep you on a safe path, steer you around obstacles and take you all the way to your destination. This is an experience that is as familiar to a blind person as holding onto someone's elbow just like I did with Paul when I came onto the stage so quick to learn easy to use and affordability is key. Most blind people can afford to have a cell phone glide is designed to cost like a cellphone 1499 for the device, a $30 subscription getting making mobility accessible to millions of people rather than just thousands. Now under the hood glide is powered by our proprietary sensible wayfinding system. This is the most advanced foundational model for set for effective navigation in complex unstructured pedestrian environments. This means that we are not only tapping into the multibillion dollar consumer blind navigation market, but we are also. But this uh we this platform will power the future in which embodied agents, intelligent guides, humanoids and other robots inhabit and share the space in which we live, work and travel. We will also empower businesses like hotels and airports and uh hospitals to make their spaces more accessible, improve the customer experience and critically reduce risk of litigation uh due to uh inaccessibility. Now, our glide has been tested by over 1000 blind people to date. And what we've done is what they tell us is they finally feel a sense of hope. They tell us that this is the most important innovation. Since the arrival of the iphone, we've collected almost a million dollars in pre orders. We've hired an amazing engineering team, folks from Amazon from irobot from symbolic and other places to deliver the product. We are raising funds. Um We are raising funds to bring the product to market in about 12 months and really position guidance to dominate the intelligent guide category. So if you're interested, please please reach out. We'll also be do doing some demos outside and I promise you that they will come a day in which you will delight in seeing someone glide past you on a sunny day, walking their daughter to kindergarten. Thank you.

Patricia Brennan, Vice President, Science Technology, Chan Zuckerberg Initiative, Tilak Mandadi, Executive Vice President, Ventures and Chief Digital, Data, Analytics, and Technology Officer, CVS Health, Shiv Rao, CEO and Founder, Abridge Moderator: Verne Kopytoff, Senior Editor, Tech, Fortune, Co-chair, Fortune Brainstorm AI

Brainstorm AI 2024: The Role of AI In Human-Powered Health 

Patricia Brennan, Chan Zuckerberg Initiative, Tilak Mandadi, CVS Health, Shiv Rao, Abridge, Verne Kopytoff, Senior Editor, Fortune

Hello, everyone. Thank you to our panelists for, for coming and talking to us today about a topic. It's very important that our lives depend on it. Uh, first off, can we perhaps start, start with you to? Um, you know, CBS is a, is a huge company doing a lot of things in A I, what is your favorite? What's your, who's your favorite child? Favorite use case? Are you sure, go ahead. Alright. Um, look, I think the, the um, anything we are doing with A I that makes our care and clinical professionals job easy is my favorite. So we have care managers that call people for next best actions and so on and so forth. We implemented, uh A I based uh, case prep that saves them 4040 percent of their time is now saved so they could spend more time with the, uh, with the patients so that those are the kind of things I like most and then Patricia to you. Uh, I, I can guess what your favorite thing is today. Yes, she has a bit of news. Um, today, I'm very excited that, uh, at CD I, we're announcing and launching um a platform to make uh kind of a one stop shop environment for biologists and ML developers to access models um of biology across uh cell data sic data, as well as imaging data. And these models are really designed to help uh biological research go faster and save time. So we can get uh to understanding disease uh quicker and, and get the cures um in a greater pace of time. Uh Sure. Could you tell us what, what a bridge does? Yeah, absolutely. So we're really focused on unburden, unburdening clinicians from clerical work so that we can bring the most important people in health care professionals and their patients closer together. So when you think about it, I'm a practicing cardiologist. When I see patients after every single encounter, I'm always mindful that I'm really serving multiple stakeholders. I have to write a clinical note for example, to give other folks on the care team. My colleagues an idea of why I fought through this differential diagnosis, why I prescribed a certain diagnostic or therapeutic. I also have to think about revenue cycle because in this country, we're not compensated for the care that we delivered. We're compensated for the care that we documented that we delivered. So these, these documents have real implications on all things related to what keeps the keeps the lights on in the hospital. And then there's the third constituent, the third stakeholder who's the most important, there's the patient and the family member. And if there's a term like transcatheter aortic boal plasty in my note, then I'm understandably gonna get epic inbox messages or phone calls asking what was that term? You never said it. I googled. It sounds scary. And so what we're focused on doing is taking from one conversation from one conversation, all the information that we need to do to do to do a kind of style transfer where we can serve all of those masters such that the clinician can just focus on the patient in front of them. And so like your, your company is a partner of, of a bridge. Uh uh how, how has it turned out? And I mean, have you encountered any wrinkles with the, with service? No, we, we work with uh Shiv and the team uh for the ambience uh uh in our uh in our clinics. It's working really well. Um It allows the doctor to pay more attention to the patient as opposed to turn around and start typing stuff, which is quite annoying. Um There, there's also um uh a I that we use in our call centers that has the same similar impact. So when you call somebody, when you call the insurance company, what have you, the most annoying thing that can happen is let me put you on hold while I pull your stuff together and 10 minutes will pass. Uh That's just that, that should not happen. So the way we are using A I in that context is Dutch stuff would come together in real time, so the agent continues to speak. So really good, good, good, positive impacts. And Patricia, I mean, health care is a very regulated field when it comes to data and privacy. I mean, uh how much harder does it make to innovate in a field in the field and, and make progress because of it? So I, you know, with uh the research that, that we do and the uh products that we're focused on, uh it's really at the sort of very early stage in the basic research. So um one of the things that uh we've really been able to do over the last several years is build up one of the largest um integrated uh data sets of single cell data, which now represents over 100 million cells. Um And those data, um we've been able to, you know, make sure that you can see, you know, where this data has come from with the studies, they link back to the original papers and whatnot. So it's really about making sure there's transparency uh to the data that underlines the models. And I hear that uh we have a new admin administration coming in and they have some ideas about A I and, and health care. What uh what have you, what are you expecting? Tla um Look, I my view is, is pretty straightforward. I think there is an arc of how A I is going to transform health care. It's an inevitable arc administrations may uh change the timing a bit, may, may go here and there some side branches. But the overall arc in my opinion is going to be in the direction of A I is going to augment clinical health care, never automate. Um A I is going to take a lot of administrative work off of the plate. A I is going to make human errors less and minimized. I think all those and, and that is going to help clinicians to get the information they need in real time. Uh That's very difficult to get otherwise. So I, I think that's gonna continue, that's going to be fine. Administrations are not. And, and so do you think that, uh you know, the Trump administration will be any better or, or worse than, than the previous one for start up? Well, I think the administration is absolutely gonna promote responsible innovation. You know, you think about it with health care in this country right now, two out of five doctors don't want to be doctors in the next 2 to 3 years, 27% of nurses don't want to be nurses in the next 12 months. So we have a public health emergency. We have to find a way to, to catch up to this emergency and technology has to be a part of the solution. And I think there's a recognition of that at every single level where I think it gets more complex is that not every use case. Um Can we, can we sprint right into? And I think that there are the lower stakes use cases like clerical work, for example, where we're simply trying to take off people's plates, the work that nobody wanted to do in the first place. And then I think there are higher stakes workflows or use cases around, for example, clinical decision support where we necessarily need to have some friction, some good friction in those workflows in order to make sure that we can deploy those technologies responsibly. So I think at this moment in time and what we've seen over the last two years is that those, those solutions, those products, those technologies that are on the safer end of that spectrum they've scaled. And I think that they'll continue to scale probably even more quickly over these coming years. And at the same time, I think we also as an industry have to start thinking about what those next innings look like even around the clinical decision support and how we can responsibly set the scaffolding for us to go after them. And, and um insurance claim denials have have reached the headlines following the murder of the United Health Care CEO um does a and I should say CBS owns uh an insurance company, does A I play a role in the insurance claim process and is it possible that A I could help uh with any the error rate of denials. First of all, it was a horrific thing that happened and sincere condolences to the family and United Health Care. Um Look what do people want from, from an insurance company. They want transparency if it, why did that decision happen? And an explanation that is in real time, not later. Um transparency to benefits. What is the coverage and transparency to cost? What what is my out of pocket cost and um an easier way to uh present a case, right? These are these are fundamentals that, that people want in a in a in a experience, healthcare experience and insurance companies uh work very hard to provide that A I is going to help immensely in all those things. Um speed up, you know, why can't we have real time preauthorizations at the time of care? The doctor is sitting there with the patient, they should be able to do a preauthorizations uh at at that time of care as an example, right? Um exposing the insurance contracts especially in the commercial space tend to be pretty complicated. So A I can help navigate that very quickly in real time to show what is covered, what's not covered and so on and so forth. So yes, I think it's an ongoing process. You know, we've been working on this for a for a while and A I is going to help even more to address those, those things, the transparency, the explainability, uh the ability to have uh a presentment of case and so on and so forth, make it easier. Uh In a few minutes, we'll go to audience questions. So please uh come up with a great one for us, please. Uh Patricia, um you know, everyone knows that health costs are, are soaring. Um Do you think that A I uh is a solution to at least some of that? So I think that um A I can help accelerate how basic research is done, for example, um or save time in uh across the entire sort of life cycle of, you know, sort of a binge to bed as it were or a binge to care application. Um And so I think there are opportunities to um you know, save the time of a researcher um by bringing more computational capabilities, more tools to bear on the work that they're doing and get the research out faster, get it to the clinician uh or the application health care um faster. Um There are some, you know, costs associated with scaling up and, and um you know, actually applying A I in, in the research process. But um I think ultimately it's about saving time. Uh Maybe we could go to questions here. Uh We have somebody uh with a paddle. Yeah, good sir. Go ahead. Um and state your name and affiliation and go ahead. Thanks by then. Um co founder and CEO of NANA, where we're a home based digital diagnostics platform. Uh sh if you touched upon this point that I thought was interesting, I think we can all agree in health care that A I will be great for workflow optimization, logistics, things like that. That's not as much as the next bar of diagnostics and therapeutics. What is the panel's opinion on some of the new aspects around that la that latter bar around validation liability. This concept of assurance labs versus FDA reduction of false positives and false negatives. Obviously, if we can corner and figure that part out, there's a real unlocking revolution in in health because curious on you guys' thoughts there, I can, I can kick things off uh maybe to, to to lock's point as well. The only currency that matters in health care is trust, the stakes can be so so high, especially in care delivery. And if trust is some combination of transparency, reliability and credibility, I think what a lot of us are starting are trying to get our arms around is like what does transparency but also reliability look like if if you're, if you're using a solution in the wild, how, how accurate is it? How reliable is it? Um And, and how do you measure that? So in our space, you know, we, we have technologies that under understand language and try to recognize the words that people are saying regardless of what language they're speaking in, even we have technologies that try to sort of a bridge the conversation, sort of distill the essence and structure that data as well. We have technologies that try to summarize as well for all the different audiences. And so for, for some of these challenges, it's pretty straightforward for us as, as, as um on our machine learning team, on our science team to sort of understand. OK. That's an F one score. We're gonna be measuring precision and recall or OK. For, for, for this, we're gonna be measuring medical term recall and um word error rate, for example, for speech recognition. But for some of these technologies, it's a little bit more challenging, for example, on summarization, how do we measure ourselves? How do we benchmark ourselves? And so 11 pursuit that we've invested a lot of time in cycles into is just being able to publish benchmarks around where we are today, being able to also give people a sense of how we assess quality ourselves within the company. And also trying to like expose these playbooks out to the broader industry so that we can all learn from each other. When I think about these assurance labs, I think to some extent, I see them sort of pulling inspiration from, from what a lot of different parties are publishing. And will we come up with one standardized sort of checklist on how to think through any given technology in any given use? Case, I think it's going to be a challenging task. And I think one of the reasons why is that the ground is moving underneath us so quickly, like these technologies are evolving so quickly that it's, it's hard to sort of take a snapshot of, of any given solution, much less problem and sort of call it done. You know, we take a lot of pride that, you know, bridge on Mondays is, should be great, but it should be even better on Tuesday, Wednesday, Thursday, Friday. We also take a lot of pride in sort of moving up the stack and expanding the scope of what we're building into other use cases. And so how do we, you know, create these standardized benchmarks for all these different technologies and all these different workflows. It's, it's tough, but I think we need some kind of like high level rubric rubric probably across the industry to just sort of guide us. My sense is that it's gonna come down to proxies for trust anyone else with a question. Go ahead. Do we have a paddle up here? Uh Lorraine Frazier at Columbia University and I happen to be the Dean of Nursing and I agree and disagree with you. I think the only currency is off of the dollar and trust hospitals, insurance companies. Let's be honest. How do you bring ethics into developing these models of A I? I would like to understand that you Patricia. Sure, I'll start with that, I mean, I think that uh ethics uh in all aspects of trust, you mentioned privacy already having transparency into uh the data actually. Um considering even before the the research uh into the model starts, what are some of the scenarios and use cases that are really trying to be enabled uh by them and bringing those to bear throughout the process. Uh benchmarks are a huge part of this. There's a I think real opportunity and gap in in the uh biomedical research field as it relates to uh sharing and uh bringing some consistency to how benchmarks um are uh made available across the research uh life cycle. Um So I think it, it really is about applying some of these core principles of uh privacy, data transparency. Um openness is another uh key aspect to this. So if you could probably take that offline, we, we're out of the time here. I wanna thank our panelists uh for a fabulous uh conversation. Thank you very.

Presenter: Ling Shao, Founder and CEO, SpectrumAi 

Brainstorm AI 2024: Improving Autism Care With Intelligence

My name is Ling Shao. I'm the founder and CEO of Spectrum A I. And I'm here to tell you a story of how an autism start up found and stumbled into a really unique business model of how to collect labeled human behavior video data. Now, at our core, we're trying to help improve the quality of care for kids in the spectrum. But in the process, we managed to find a way where we can get paid to collect this labeled human behavior data that has implications far beyond autism and even far beyond health care and in other industries, for example, uh education. But let me start at the beginning of my story. So I'm a mom of four boys who are all on the autism spectrum. And for those of you who are not familiar with autism, it is uh in essence uh a condition where kids have really significant challenges, learning language and how to socialize with others. And even after they learn language, sometimes our kids are uh labeled as acting like robots because their uh their physical kind of how they portray themselves physically does not necessarily line up with how they want to communicate and sometimes they speak with a really flat intonation. Now, full disclosure, uh I went to MIT as a nuclear engineering undergrad. So I'm very familiar with being socially awkward, which is makes this a little bit really awkward right now. Um So I thought my kids were going to follow in my footsteps, but as it turns out, they had a real medical condition and um because they all started out nonverbal, the only treatment covered by insurance was called applied behavior analysis. And what that means at its core is basically you have clinicians using behavior data of how your kids respond to try to figure out how to teach them how they learn what's the right education protocol to apply to them. And now my little guys uh they all learn through what I call pattern recognition, which means that when we are teaching our youngest son, the word for ball, we showed him a picture of an orange plastic ball and said that's ball. But he could not generalize ball to a picture of a green ball or purple ball or even the 3d actual object of that orange plastic ball. We had to show him tons of different photos and even 3d images or 3d objects to say, what are the possible things that could correlate with this word? Does that sound like something else we train? Right? So I had this light bulb moment of oh my gosh, the way we train our kids, how to use language and socialize with others has a really close parallel with how we train our A I and ML models. But let's hold on to that light bulb moment because I had a more pressing need. I had to figure out how do I find the best clinicians with the best clinical protocols to treat my kids? And as it turns out, as you heard in the intro, I am not alone, the incidence of autism diagnosis has increased over 300% in the last 20 years. In the last five years, we now have mandatory insurance coverage in all 50 states. So I started Spectrum A I as a way to pay it forward and help all families like mine be able to access high quality clinical care. The way I was going to do that is to marry the insurance reimbursement to pay higher for those clinicians who actually deliver better outcomes. Now that sounds super obvious and super simple and for the healthcare folks room that's called die based care, but it is incredibly difficult to pull off because we don't have objective data that the health insurance companies can audit to prove if these kids are getting any better. So the way we solved that at Spectrum A I was we built an electronic medical record, the clinical documentation portion of it to enable therapists to collect data during their sessions. And more importantly, they were labeling the video of the actual sessions. So we've recorded the video of each session and as they were taking data on how the kids were behaving, they were actually labeling our video. Now, this is important for a couple of different reasons. One, it helps clinicians collect data easier and more time efficient. So they're spending more time with the kids and less time creating reports. But more importantly, it gave health insurance company an objective bank of data sources that they could use to compare across aggregate physicians. And then because they could audit the system, they can actually reimburse based on the outcomes the clinicians were actually delivering. So I'm really happy to say that three years into our start up this year, we had the first national health insurance company launch an initiative where they paid providers who used our software, a higher reimbursement rate because they had access to this objective data. And in Q one, we're actually the launch with two large Bay area employers. I'm sure you guys have all heard of um to allow their families to access higher quality providers as part of their employee benefits. So in the near term, I think it's a pretty easy jump for you guys to see how we're going to use this data to actually start to automate labeling of some of these more common behaviors, right? That again saves the clinicians more time. But in the longer term, as we collect a larger and larger data set of accurately labeled human behavior data because remember it's being labeled by clinicians who are literally trained to label human behavior data. The use cases of this data goes far beyond just improving autism care. So for example, education, imagine a world where we can measure and figure out which kids actually learn better using which teaching techniques. I started my career out as an army nurse and I can definitely see a military use case for this. Imagine the military using this to assess behavior of their troops as we think about discharging them for battlefield readiness. Now, I'm going to end with bragging a little bit about our tech because uh here is actually a footage of my youngest, remember the kid who couldn't generalize the word ball. Um This is actually a screenshot of the interface that the supervisors of the front line staff used to see the video that's been labeled during his session and allows them to both edit and annotate review and annotate what is actually being labeled so they can make the data source even more accurate. And two, this is the data that insurance companies can review and audit to see what progress the kids are actually making so they can validate why they're making these higher reimbursements. And if you zoom in, you can see that uh while we made this system to really help label the behaviors of these kids to figure out which protocols would help them best this label data that we have generated and get paid to generate is really what we think could be the best value of our company. And it's something that we stumbled into but believe have far-reaching applications beyond just helping these kids. So we started using this data to help kids who um acted like robots how to communicate and socialize with other, with their peers. But nothing stops us from using this data to literally teach actual robots how to interact with each of us. So for any of you who feel like this data could be used for something you're working on. I hope you will join me on my journey to use this data. Thank you.

Prem Akkaraju, Chief Executive Officer, Stability AI, Dana Settle, Co-founder and Managing Partner, Greycroft Moderator: Allie Garfinkle, Senior Reporter, Finance, Fortune; Co-chair, Fortune Brainstorm AI


Brainstorm AI 2024: A New Lens on Stability

Prem Akkaraju, Chief Executive Officer, Stability AI, Dana Settle, Co-founder and Managing Partner, Greycroft Moderator: Allie Garfinkle, Senior Reporter, Finance, Fortune; Co-chair, Fortune Brainstorm AI

Prem Akkaraju, Chief Executive Officer, Stability AI, Dana Settle, Co-founder and Managing Partner, Greycroft Moderator: Allie Garfinkle, Senior Reporter, Fortune; Co-chair, Fortune Brainstorm AI

Frame Dana. I am so excited to be here with you in part because this was known on my Bingo card for 2024 talking about a turnaround for this company less than eight months ago, stability was in a rapid decline. But here you are, you both stepped in and decided you really wanted to take it on why it wasn't on my bingo card either actually. But no one, no, eight or nine months. Yeah, it's no one's Bingo's Card. I think that um well, it, it was really because the, the, the technology and the team were too good and uh I've been partners with Sean Parker for over a decade and we have just did a very similar uh play at Weta Digital where we created an application layer for film and TV. We made Avatar one Avatar two big movies at that with that company. And we were after and that was we sold that company in November 21. And after that, we're like, ok, we should, we should probably be looking at the A I version of what we just did and how could we use A I to make film and TV, production gaming, as well as marketing, advertising, more efficient. And then this opportunity came about and I got heavily involved with Dana um on this like earlier this year. And um you know, this is where a lot of the outweighed returns live is when you're willing to go in and do the very hard work to kind of turn it around. And we did that very quickly and in opportunistic situations. And I think, um I'm, I'm fortunate to be a part of the uh the partnership now, the tram brick along with many other partners. Uh But, you know, like Prime, that is where the outsized, you know, returns live is actually in jumping into opportunities that, that other people don't always see. Um, and especially in an area like this, that's moving so quickly and where the technology here, like Prime said was just, it was, it was too good to let go. Dana. Dana bring the, sorry to interrupt the, the uh um she brings up a very good point. Like opportunities are things that only a handful of people see if we all saw them, they wouldn't be opportunities. And so that, that's kind of the definition and I think that we saw it and everything that we've done for this. How many months has it been now? It's been six months, six months, last six years. How long a time? Exactly. It's like 100 and 52 years in like real time. But, uh, we, we've seen extraordinary results in such a short period of time but, and it was really, it's, it's good to see that, you know, our assumptions really coming through and what was the financial state of the company when the two of you came in? What were we dealing with here? She's the, yeah, I mean, what are we looking at the financial state of the, they needed funding and, and that was, you know, where we had to really move quickly on analyzing the opportunity. And so we moved very quickly, but so deeply because we wanted to make sure that it was what we thought it was. And, you know, like Prime said, it's proven out to be, I mean, e everything we thought and more and I think in particular for us, it was really backing an incredible leadership team in Prime and Sean Parker to, you know, come in and, and lead what is essentially doing what they had done just a few years earlier in the South Space. I mean, you said they needed money. What for how bad was it was it because the reporting at the time, my colleague, Jessica Matthews and I spent a lot of time reporting on the story and the, the narrative was that it, it stability was a dust store. Well, I mean, look all A I companies require a, a lot of capital. I mean, that's no surprise. I think you can read it. Anywhere. Um But you know, I, I think what, what happens when you come in with a really incredible experienced management team, leadership team is they know how to get in. And this is prime, I mean, Prime really was working with the team um at stability even before we close the transaction and, and uh and, and helping to identify, you know, exactly where were the areas that they needed to focus on for this opportunity that we saw in visual media. I would say that people forget that Apple was 90 days from bankruptcy in 1997. Actually, I think they were probably closer to bankruptcy than actually, you know, this company was. And again, it all comes down to leadership and you saw what happens when leadership comes in with a very succinct plan, a product vision and the ability to, to attract capital obviously. But, you know, the clean up, we have now a clean balance sheet. We have no debt, we have nothing and to negotiate out of those deals or within those partnerships. It was no miracle by my part, even though I led the transaction. Give me an example of what we're, how much of a mirror. Well, I mean, it, it was, again, this wasn't on anybody's bingo card again. Well, I'll take, I'll actually give the credit. Well, I appreciate that. Thank you. But um while I led that effort, it was, we, we cut deals in the best way. That you cut deals, which is um alignment of interest. And everyone on the other side of the table that I had to deal with to clean it up. And then we got there was saw that this was essentially too important to the ecosystem to fail. Stable diffusion. Represented 80% of all the every single image that everyone in this audience saw in 2023. Just last year, 80% of them came from stable diffusion that's gigantic. And when you think about what that implies with a 5 million person active community that works on our models every single day that's growing every single day. That's a gigantic ecosystem that the infrastructure of this industry could not take a hit on. So I think it was part, it was, it was, it was the alignment of interest with those parties and um uh you know, and our determination to make sure that that was articulated. And, but I mean, the reason I say you should take a lot of credit for it is, I mean, really bringing everybody along at a time like that is, I mean, that is true, hardship and that is hard and it's in doing it too when, you know, I mean, frankly, you weren't even ceo the company. And so, I mean, I think when you think about how difficult it is to lead a company during, you know, tumultuous times and during super high growth and, and change, that's hard when you are the CEO but, but, but what about when you're actually just coming in and, and sort of coalescing, you know, all of the different groups from, you know, investors, employees to partners, et cetera. And I think that, you know, that's what gave me greater and greater conviction at every step along the way, was just seeing how prime actually executed through that. Well, controversial past leadership has left the company but a lot of the early investors in the company, including including Co two and Lightspeed, they left and then they came back. Yeah. How did that go? How did that happen? Is that incredible? I'm very, I'll be honest, I'm surprised that I think that says it all, you know, and I think they really re underwrote to this leadership team but also understood deeply the opportunity. Co two left the board, they didn't invest anymore. Same with Lightspeed, same with sound ventures. So not only did all of three of them reinvest into this round, um their pro rata that they were allowed on only the best we did it, we did it as a rights offering in, in, in some cases even more than their pro rata. And not only did they did that but coach who actually rejoined the board and is on our board again now too. So it was uh it is extraordinary because I think they saw that progress being made and kind of the positioning of re establishing the leadership position. The reason that everybody is interested in the company is because they were so successful. If they weren't successful, you wouldn't have these, like, no one would care. What are those conversations like that? What was the vision that you, that you sold them on? Well, again, it was alignment of interest, I think in those, you mean for the other investors, it was really well there, it was really my pre previous success. And yeah, yeah. And I was gonna say, I mean, I think one of the really exciting opportunities here that, that we saw is a new investor coming in was essentially um you know, where we think that there is great opportunity in general A I right now is, is really when you're sort of coupling the models with the application there. And I think these models were phenomenal are phenomenal continue to be and then having a, a leadership team that's so deeply throwing stuff around under understands the the do exactly heads up um is really an extraordinary opportunity. And I, and I think that's again why I mean for us as new investors coming in and, and we are based in Los Angeles, New York and Silicon Valley. But we, we really understand the media ecosystem and knew how deeply prime understood that and, and was connected in that ecosystem. That's really that, that transition because we were, you know, the first evolution of A I was really the foundation of the frontier models. And then it was quickly seeing this base, um because people always talk about the application layer and it is true that value is there. But being able to understand and fine tune models all the way into a core model is extremely important for you to set up success on the application layer. So I don't think, well, you know, some people believe, uh you know, foundational models are, is raced to the bottom. I I, you know, I uh I tend to agree as a stand alone. Um We're really that application, the applied research and application layer, which is a race to the top, but that's a race to the top that we can have an advantage winning when we understand the entire entire software uh stack all the way down to the chip. Now, I want to talk a little bit more about the board coach. He's back. You da are on the board. James Cameron is also on the board. Yeah. Who here has heard of James, I actually who doesn't know who James Cameron is? Avatar? Avatar? Yeah, just a just stream avatar. Um How did that happen? What does he bring to the board? So I worked. So when I was CEO of Weta Digital, we made Avatar two made every frame of that movie with, with uh James Cameron. And so um when I built the application layer at WETA and eventually sold it to unity, uh you know, he had a front row seat at everything that we were, I was doing and uh with me and the engineering team. And so when this came about, um I was in constant dialogue with, with uh with, with James. And, um and in fact, pulled him in early in the year before I joined the CEO and I said, hey, I think this actually could be. And he said what he said, he agreed. He said he agreed, you mean about that? This could be, this has the potential. Of course. Well, I don't think he would have invested or joined the board if he didn't, I assume. Yeah, that makes sense. That track. But even back then before you joined, before it was before this full turnaround happened. Yeah. Well, no, it was definitely, it would mean he agreed, it would be hilarious if he did join the board without thinking about that. No, no, no, no, no, no. Well, there was no board for him to join at the time, right? I wasn't even the CEO of the company. So what I mean was he agreed was, was the question that I posed to him, which was, hey, I think this is worth looking at and like, let's really dig deep into this. And like Dana was saying, we dug very, very, very deep into the talent pool. Like we didn't, we didn't invest and I didn't join the company until I met every single person at the company myself. Um and Sean and, you know, we dug very deep in the talent pool was so depth, was so deep. I was so surprised because there's a lot of also like in no offense, the media kind of, you know, over exaggerates in both directions many times. And then so II I was reading but she's laughing because it's true. Um But I was so shocked because this was because there were some departures and there's other, you know, spinoffs that came off of it. But we were so um so happy to see that not only did the depth of talent still stay there. And there was over 90 people within the core and applied research team, but they had two other things that were, that you that are beyond just their genius and they were dedicated and they were very passionate about stable diffusion. And I found, and I, I told this and just recently, um during all that, you know, that was what feels like years ago, um was months ago when I was meeting everybody and when we close the deal, there was somebody on the applied research team that said, we knew you were coming. This is even before the actual deal was announced and closed. And they said, and what's that? Yeah, I mean, it's an amazing statement to make meaning, not me, but like, you know, somebody was coming that was experienced leadership that could bring the vision that could bring the capital to the table and clean up the company because it was just, it was too important. Well, they might not have imagined that it was gonna be you, Sean Parker and James Cameron. But to be fair, who could have imagined that again? Everybody's surprised Bingo card. What is stability do to today? And what is the opportunity? It still it down for us here and now. So our vision is we're doing two fold is one is we're exploiting the commercial opportunities, significant commercial opportunities with our enterprise uh licensing and API revenues. And the second is we are now building the first, first full stack A I production pipeline for professional vi visual media content, professional video media content, defined as film and television, obviously gaming and marketing and advertising. What about you? The opportunity? That's where, where, where does, where does this land for you? What's the end game? It's, I mean, it's amazing because it really is sort of the thesis that we had laid out at the beginning, which is that, you know, there is a huge opportunity on the creative and production side of media to leverage A I in the creative community. And really, I mean, you know, if you think about it, we have this incredible product marketing genius and James Cameron on the board deeply involved and having that combined with this technology is so unique and being able to then apply it to those three verticals verticals that, you know, we at gray crop know very well. Um Having been involved with, you know, companies like Scope Le, uh which is a large gaming company which was sold uh last year, um, you know, early days of youtube with Maker Studios off the TV. So this is an area that we're in, you know, mountain on the advertising side. We just think there's such tremendous opportunity. Um Sorry. Yeah. No, no said perfectly, but I, I actually don't know how to do it any other way. Um I think it's pretty because I did it at Weta. That way, we were, there was no space between the art and the science. Uh We worked with some of the biggest directors in the world. We made the last 10 Marvel movies and, you know, obviously avatar one and two and we worked very, very closely with the directors. And uh that was such a great um experience for me because people ask me all the time after we sold, after I sold the, the pipeline to unity. It's like, how did you know what products to build? How did you who designed your product pipeline? I was like, it's easy. I just asked the directors what they wanted. And so it, and it, it's true. So it, it became the product strategy, the product road map, the go to market strategy, everything and also the Q A everything is sort of built into one. And uh the confusing thing I find is how am I supposed to build uh artist tools without an artist as a partner at the highest level of the company? So that's just it, it's my, it's my muscle memory. Now, let's imagine for the sake of argument, we're sitting here in a year, us and a couple and a couple of our nearest and dearest. What has stability achieved? What does stability look like a year from now? Like robust managed service on our um which is growing at multiple, multiple, the, the, the, the head of marketing told me I couldn't tell you how uh in front of everybody, how much we've grown. But it's been extraordinary. Exactly. I would like to, I would like to because it's a gigantic growth. I know. I know. I blame you. Yeah, exactly. I'll tell you in a year, I think, let me then on here you can ask her the um uh so she wouldn't let me talk, but it's, it's incredible. It's triple digit growth and the uh and so I think we're gonna continue to see that and I think that we're gonna be seeing um several breakthroughs in individual shot elements. We have a very different approach. We have a professional approach to content creation. We don't have one text to video model that has one output that combines like a gazillion, different workflows into one that's actually not how professional content is created, professional, created in shot elements and then they're composited together, by the way, this is where he's going to go really, really deep because he doesn't understand this thing at all. Models. Now, if you want the, we're sadly at time. But that's, that's, that's for, it's ok. I can go over. I mean, they all came and way more interesting. Thank you and Dana so much. That is all the time we have. I wish we could. Thank you so much.

Pegah Ebrahimi, Co-founder and Managing Partner, FPV Ventures, Rajat Taneja, President, Technology, Visa Moderator: Terri Burns, Founder and General Partner, Type Capital

Brainstorm AI 2024: The Greatest Financial Instrument of All

OK, Pega Raja really excited to have you both here today, Raja, maybe we'll start with you. Visa, of course is at the forefront of leveraging A I for risk prediction. Can you tell us a little bit more about this? Tell us how Visa approaches the challenges of scaling A I solutions to detect and mitigate A I risk globally. Absolutely. Uh Terry, you know, Visa is the OG of SAS companies and it was one of the original SAS platforms and lean in to A I back in 91. So core to Visa managing about a billion transactions a day, connecting 4.6 billion credentials to over 100 and 50 million merchants is the ability to do it in a trustworthy manner in a secure manner, managing the payment, security and cybersecurity. So our fraud model have been at the core of what we do with A I for over 30 years. And we went from good old fashioned A I to deep learning to neural networks to all the predictive A I pieces to now leaning in to generative A I helping us with models with the problem of Cold Start for something like a real time payment network to figure out if there is any problem going on with account that are being enumerated. Uh to see if bad guys are trying to see if cards are legitimate before they use it for nefarious activities real time uh management of this. And it's all about investing in the data platform, spent about $3.5 billion in the last decade to create the data platform for the future. And it's all coming together. Now, I want talk more about how you and your team have really embraced this culture. But but first, Pega, you have had a fascinating journey from CIO to investor. Can you tell us a little bit more about trends that you're seeing in A I and specifically, what do you think could have the biggest impact on enterprise transformation or even cio decision making? I look, I was I was a cio during the first mobile when mobile wave was coming. And it's like it's super interesting because you see, you know, everyone's like A I is so different than before. And yes, there's a lot of things that are different in the pace that it's growing. But in a lot of ways, history doesn't repeat itself, but it rhymes and it reminds me so much of when mobile first came and everyone was, there's all these mobile companies that existed, right? And for the most part, nobody says a mobile company anymore, every single company that came was on cloud is now born in the cloud and using mobile and, and it's part of being company. And I would say that's the same, you see it happening with A I at the beginning, it was like A I company A I company. And I think now there's definitely going to be foundational infrastructure A I company, but for a lot of the real great companies that are going to form, it's going to be how are companies starting and taking advantage of the fact that these inherent advantages exist because of A I and to build something that they couldn't have built before, right? Similar to when Uber, nobody thinks of Uber as a mobile company, but it wouldn't have existed without mobile if you don't have your phone walking around, um you couldn't really do that business model. So that's the part that's really exciting. It's how do you take things that people just assumed would have to be services, businesses, things that people assume would have to um you know, depend on so many people to get it done or it couldn't have been done because of the scale it needed. Those are the parts that are really exciting. That will be the next wave of A I that I think are, are gonna be exciting start ups. Let's talk fintech. This is a question for both of you. Do you believe that fraud prevention presents the most significant application of A I in fintech today. And if not, what is something that is commonly overlooked that you each are really excited about? Look, I think fraud is table stakes managing it, managing it well, staying ahead of the bad guys uh because they're very proficient, the same tools are available to them. If cybercrime was its own country in 22 it would be a $7 trillion economy growing at 32%. So it's very profitable. It's very big and so fraud and payment security and cybersecurity are existential. We have so many attacks on our network millions that you have to stay ahead of. And A I is a very important tool there. But I think there is the other side of A I. How do you use A I to re invent and rethink commerce? If you think about it, the last big major change in commerce was with ecommerce and then mobile and continue to break back. Years of shopping in the storefront or online. Now comes the opportunity to move from the hunt and peck era of shopping where you are still searching for things, you're analyzing it, you're trying to figure out what to buy, then going to the buying mechanisms to what we are calling the self driving commerce era where A I takes away some of the mundane aspects of discovery and of helping you with warranty management. Price matches registration of products and it brings the joys of commerce back to you in ways that you're discovering things that you would like to see whether it's products or travel or, or art and then helping you really enjoy and appreciate the pleasures of shopping. I think that's where a lot of the innovation is going to happen in addition to the basic foundation of core infrastructure, fraud and security. Yeah. And I always loved it like as a, as, you know, a cio or it was always nobody s companies had a hard time selling to financials because it time if you think about it, all the things in financials were on prem and it was just really difficult and I would say what's really exciting about this time around and how it took financial companies in a lot of ways, so much longer to get advantage of the digital transformation trends because they had to really take their time, make sure their data was secure, do it in a much longer time. Horizon right now, financial companies, I'm seeing move way faster in A I than some of the other industries, right? Because it almost was, they had to get set up, they had to care about data, they had care about some of these regulations and some of those parts make it easier in certain use cases for them to adopt faster. And I was always, when I went into tech, I was at Cisco and it was always interesting how everyone thinks of financial companies as old stodgy companies. No. No offense to us. Right. And in so many ways financial companies depended on such razor thin margins that in so many areas they had to be super efficient. Yes, they weren't the early adopters of technology, but efficiency is so core to finance organizations. And I think you see it now some of the financial visa other, they're moving so much faster on thinking where are the, where are the areas where we could actually take advantage of some of these um technologies for margin expansion for what services do we have on the fraud side, on customer service on financial advisors? I think that's incredibly interesting. You look at, you know, there's going to be 80 trillion of wealth transfer happening from baby boomers in the next few decades. And so many of the, you know, only one person and has access to financial advisors because it's just not a scalable by person business, but so many people that are used to technology and all those things I think will get advantages of that that they couldn't have before. And I also think there's just, there's, there's gonna be a myriad of other stuff that comes from building efficiencies in the orgs to all to, to actually consumers getting the benefits of it. Yeah, I think it's a really interesting point what uh in recent history might have been seen as kind of old school is now really at the cutting edge. And I think that both of you uh alluded to that in, in your responses in a couple of minutes, I'm gonna turn it over to the audience. So if anyone has questions, start getting those ready. But first, um I wanna talk a little bit about A I transparency. So there's this ongoing debate with A I models, particularly open source versus closed source. I know this is something I think about a lot as an early stage investor. What do you each think is the right approach for early stage companies? Well, I mean, I'll, I'll talk high level and then we can talk about what specifically they're doing. I think there's no one formula. I think it's so early that everyone's experimenting with a lot of different ones. But I do some of the things that are saying parallel look, I remember when the cloud migration was going and people were like, oh, do I choose Aws or Azure or do I don't want to get stuck on this cloud? Now? It's even fundamentally even I would say more exon because people are forget cloud lock and needing a few years and a bunch of engineering resources to migrate. If you have like model lock, you got seven years in and the prices change or the the dynamics change. And now you have all this investment that you did on really bringing your model up. So forget model lock. I think that um cloud lock, model lock is even becoming more key and So because of that, I think Cio S have saw that game played out. So they're just going at it more eyes wide open this time and thinking, I don't want to be stuck if something changes, if open gets closed, which is happening to some companies, right? And if things like that happen, and I think that's, that's the learning from that cycle have made Cio S just want a lot of different opportunities and knowing how to, how to be more in control and not stuck on something. And you see it at least that you guys are doing it. I think so well. Said Pega, you know, I think there are pros and cons to open and closed and you describe it really well with the M lock. I think the core is for there to be radical transparency. A key pillar of the model is the data. You have the algorithms and you have the GPUS, but it's really about the data and these are general purpose models that are built and trained on massive amounts of data. And we know the data out there has got noise in it. There is a data that is error prone, there is data that is just blatantly wrong. And so as more, whether it's an open model or a closed model as they are building it to explain with transparency, what an explainability, how they use the model, how the algorithm created fairness, consent, privacy, copyright infringement issues is super important. And for companies can I interrupt you though? Can you have true radical transparency with a closed model? It depends, well, it's difficult, right? Because it depends on what they are saying and the community is not able to go in and therefore you need to think about the regulations terry that come in and help create a framework for understanding of what happened in the model so that the practitioners can use correctly. If models are going to be transformed in medical and health care and diagnostic tools or finding new drugs, you need to know that right. And therefore, I think the role that regulators will play in creating that framework uh use based um interoperable because models don't know boundaries of countries. So you have to have global regulations that allow us to have guardrails and think through what went into it and how it can be tuned uh and how it can involve. So they are helpful and they are able to do what they are supposed to do. Yeah, I think that's fair. I think my personal bias is towards open source, but I think you're exactly right. There are very clear use cases where for data protection or sensitive situation, closed source, but with still some transparency makes sense. So right now, we have an ensemble, we have a number of different models at visa and we have abstraction layers on top of it much for the ways that pick up was saying but not, not getting a model lock in. Yeah, for sure. All right. Let me turn it over to the audience. Let's see if there are any questions. Ok. It looks like there is a hand raised here. We'll give it one second for our Mike runner. Please do say your name. Hi, I'm Maria from Freestyle Capital. I'm curious how you guys think about agentic payments evolving. Um I know that Visa has been thinking about it. I'm sure everyone has been. But do you think that it will evolve on existing infrastructure or do you think that a new infrastructure will need to be created for it to work at scale? Look, I think payments as a part of commerce is incredible, right? So you need high security, you need it to be seamless, you need it to adhere to all the regulations and compliance of every country. And as you move into an agentic model where a agent is conducting commerce on behalf of a consumer, it has to have all those attributes and explainability and management of compliance, but it will require new infrastructure because you're running models, you're in the models are doing inferences, the models are conducting transactions. So it will require digital investment in infrastructure. For sure. You sort of mentioned earlier. Uh Raja, you also mentioned this pe a regulation and so I wanna kind of double click on on this idea. How can companies balance regulatory compliance with the rapid rate of, of A I development. How do you both think about that? So, I mean, I think one of the look and we're talking about financials, which is interesting because it's been regulated for so long, which is why it's been so hard to sell to it. And you need to have a lot of data management and all these aspects. I think what's gonna be interesting is that for, for regulation is usually a little bit delayed. I mean, it's from how fast companies wanna go. And I do think that um on for most folks that are trying to enable A I it's hard for them. If you're a large company that has a high brand risk, it's hard for you to be at the forefront and then get pulled back. So what's happening is I'm seeing a lot of companies go in the areas when they know they're not going to be necessarily hitting some of those pieces. Like can you pick low hanging fruit that you can drive a ton of efficiency versus trying to go for the moon and trying to go for something that has a really high risk. And I think that's what you're seeing. Some of the pullback companies where at the beginning, there's this crazy excitement about, I wanna have po CS of every single use case on using A I and taking it back and saying, OK, what are the areas where we can do it? And we can have some risk exposure. It's not gonna be, you know, existential for the company and where are those areas that we wanna go? And I think that's what you're gonna see the next, you know, few years until the regulation catches up because I think people kind of are like, I don don't know if it's worth it to take that leap and, and have some existential risk, but there's so, I think you and not all use cases are the same. Like if you're touching personal data versus if you're having an FA Q, that's way more, you know, way more robust. That's, that's totally different areas. I think as a risk, you look, I think you have to think about the spirit of any regulation, right? If a safe, secure, trustworthy models, use of data explainability, fairness. So we have a set of guiding principles that are called data use policies in our company and we adhere to them very closely. So as regulations evolve, whether it is the EU Act of 24 whether it is the uh executive order in the US, uh the year before that or its new toolkits coming from Singapore or from the African Union. I think we just have to stay with the spirit of what you're doing and collaborate with regulators and academia to create the right framework. But if you are open and fair and transparent and explain things, I think you will be always safer with your models. Um In our last minute and a half here, you both manage teams, you have lots of stakeholders and clients and start ups that you work with. How do you all spot true value and opportunity from hype? As we know, there's a lot of hype here. What are both of your really quickly frameworks for how you think about that? Uh I think, I think the number one thing is don't you, you can always use the hype. But to me, um just the best companies don't talk about A I, they don't say, oh, we're an A I company. They say here's the big problem we're solving and here's the unique insight on how we're actually solving it. And I think that's at the end of the day, that's where the huge companies will form. We're solving real problems. You can obviously use the advantages of A I, you, you should because you're a start up and you cannot a lot. But I think those are the best ones and I think there's a lot of opportunities in the services or look at large companies. They have massive services that they have already figured out how to extend you outwards. And I think there's going to be a huge number of companies formed there. Um I think, you know exactly what you said, move away from powerpoint and go to code and let the code talk to you and utilize it and take it for a test drive and if you're helping it write code software development, use it to write software development and see how it compares or contrasts. At Visa. We give access to every single employee access to the models. Back in 23 we had been using it uh years before that and had patents and papers and so we knew that there was great potential and, but the real rubber meets the road when you use it. Absolutely. Tag Ja. Thank you so much for your time.

Rohit Prasad, Senior Vice President and Head Scientist, Artificial General Intelligence, Amazon Interviewer: Jeremy Kahn, Fortune

Brainstorm AI 2024: The Everything Agent

Ravi. Thanks for being here with us. It's great to be here. I'm missing all the bad weather on the East coast. So I'm excited to be here. Fantastic. So big news out of out of re invent last week was Nova, as Andrew said, um you know, you finally have a family of large language models out there. I guess a lot of people are wondering though what took you so long? I thought it was very fast. You know, uh I think uh first, very excited for what we did with Amazon Nova. It's our next generation of foundation models that we introduced to the world last week. Uh Let me take you a bit behind the scenes of uh uh where I sit, I have the words best to thousands of generative A I applications being built by our internal teams. And we are noticing three very important things. One was choice and importance of selection. So if you look at Amazon Nova Foundation Model family, it's not just one model, we actually introduced six models. Five of them are generally available as of last week and they start with one of the models which is a text only model, which is called micro, which by which you mean that you can take any text, input and output text. And then there are three models of varying increases in size and intelligence that are taking multimodal input as an image, text and video and generating text. And then you have two additional models that are using taking text as input or images and generating images and videos. Those are called Nova canvas and novare. And the reason we did that because we found internally, people are using very different models and they had very different needs for their applications. So that goes down to the selection of what you need to do to build. And then secondly, as people move from experimentation to large scale loads, we were finding that cost is becoming increasingly important. So these models are 75% less expensive than their counterpart best performing models on bedrock, which is our service where you have many different model providers available with their foundation models. So clearly cost was big deal and so good to see that the team did massive invention to bring costs down. And lastly, as you know, with these foundation models, it's not just about the model's capability or cost, but if delivered a very good user experience and after time for the user experience factors like latency matters. So all these models on the ones which take uh image video and text as input and output text. They are lightning fast. Uh So those were the things that we worked on and I'm very excited about the future and we are looking to the next generation of Nova as well now. Great. I think with all these models, one of the big questions people have is is data, you know, where did you, where did the data come from? That that was used to train Nova? Yeah, we don't disclose the specifics of the data, but it's a factor of many factors. One is sources, of course the publicly available data. And second, we have a lot of proprietary data and we also license data from many providers. Yeah, that's interesting licensing seems like it's becoming more of a norm in the building of these models now whereas it may not have been in the past. Yeah, I think you want to, to trust with the publishers and everything is super important and we want to make sure that you are using the data the right way and all that is super critical. And I think we, we are very excited about some of the partnerships we have made on the data site as well. So now Nova is out, but a lot of people are still waiting for an update to what is maybe Amazon's sort of best known product when it comes to A I, which is the Alexa Digital Assistant. I know you guys said there would be a new Alexa coming. But you know, when is it actually going to arrive? Yeah. As you know, Jeremy Alexa is near and dear to me, I joined Amazon more than a decade back to build Alexa and it's been a very fulfilling experience. Uh We are right now in the midst of reengineering Alexa with the latest of state of the art foundation models so that it can be the world's best personal assistant. We are going to be true to that. No T A that we set ahead and a lot of people coughed at us when we were looking at that, but that's the reality. You can have your world's best personal assistant available to everyone. And what I see in our labs is super exciting. So all I can say today is stay tuned. Uh Jimmy there, no, no timeline, not, no timeline right now. Just stay tuned. OK. Well, we'll stay tuned. Um And I want to come to the audience for questions in a minute. So please think of your, your questions. Um One of the things with Nova, I mean, you built that those, that family of models in house, but you also have this close partnership with anthropic. You made investments in them, you have a strategic relationship with them and I know they've been helping you on, on some things. Um You know, why in some ways, why build your own own family of models if you have that relationship, philanthropic and they have very capable models of their own. Yeah. First, we love Anthropic. The partnership has been great and I'll go back and reinforce what, where I started, what, what we saw was happening internally. Customers want choice. We have seen that many times in our internal Amazon stores. If you look at how many different selection we have there and same in a, there's not just one solution, there are multiple ones, same. It's turning out to be true in the foundation models. Our customers do want choice and the partnership with Anthropic and what we did with Amazon Nova is just accentuating the the choices available for our customers. Um So the the relationship Amazon has with Anthropic is somewhat similar to a relationship. You know, Microsoft has with open A I, there's been some reports in that relationship between Microsoft and Open eye that there's some tension. What's the relationship like with Anthropic? And is there any, are there tensions in that relationship where they want more compute out of Amazon than you guys can give them or you're not getting what you feel like you need from them in terms of innovation. On the model side, I think as I said, we love the partnership. I think nothing more to state here that the partnerships are going great. And you saw some of that at reinvented last week, right? Um There's a lot of discussion recently about progress in large language models hitting a wall. What's your view on that? I think that's a very debatable question in terms of what wall is coming. I think we have to be very careful about what we have seen in A I, every time, every time we come close to a wall, there's a new invention, I think we'll keep finding that if you see what is happening with pre training, which is where you build a token predictor and post training, where you're giving the model a lot of different tasks. Now it's shifting to inference. We are spending a lot more time as an inference. And at the same time as I mentioned with Noah, we are making inference cheaper, which means I think the scaling is curve is changing from different stages of training to inference. And we'll continue to see that I'm very optimistic that in A I every sometimes every decade, sometimes every five years, every three years, you continue to ship the curve up on performance. And I think that's going to be true again in this coming years. Are you working on a reasoning model that would use your test time inference the same way as the one model from, from open A I does? Again, I won't go to the specifics of it like in terms of the many different ways, what we are really building is for A I to be useful in real world, practical applications. And a lot of that requires reasoning of many different kinds, not just mathematical reasoning, but more utilitarian reasoning as well where you're trying to do tasks in the real world. And that's where if you look at A I agents, you have to get to the point where they're not just solving a coding problem or math problem, but they are calling the Uber for you or making a reservation for you, which are not really high on intellectual proficiency but require a lot of world knowledge to do it. Right? So that's where the feel is evolving. And is that where Amazon is heading towards agents and making agents available? Yeah, I think if you saw like what we talked about with our Rufus, which is a shopping assistant with Alexa, which I think of as a super agent agent because it does many different things. That's where the word is going to go, right? I want to get a question from the audience. Please raise your hand if you have a question, if not, I've got lots of questions for. So keep going. There's a woman down here, wait till we can get a, a mic to you. If we can get a mic handler, just a woman in red, just down here. Bye bye. And if you could please state your name when you stand up and where you're from, that would be great. Thank you. Hi, how are you? My name is Lorenza and um I work currently for the Press uh at the Economist and uh yeah, I would love to. Well, we know that Jeff Bezos is back working uh at uh Amazon full time. And uh yeah, I would love to, to know whether you've had uh you know, work with him in, in certain projects. And what is his vision? Has he mentioned? Any specific vision uh that you can share with us on his mind? I would say you should ask that to Jeff. But yes, he's very involved, as he said, uh I love the partnership with Jeff. I've worked with him early days of Alexa too and he's still very much connected with A I elements of uh where I spend a lot of time and I get to see him. So very fortunate with that partnership and mentorship. OK. Um I want to ask a question about A I's energy demands. We had an interesting session this morning at a breakfast on this topic. How is Amazon looking at, at the huge energy demands that particularly these foundation models seem to require? Yeah, I personally on my front as well. I want a I to be sustainable and I love that at Amazon, especially with Aws. What we are doing is we are looking at the end to end life cycle of sustainable A I. And it starts with, I think our chips. One of the main things where we have our mlx and chips like trainum is essentially to drive better price performance so that you are able to build these foundation models in a much, much better and sustainable fashion. And specifically, if you look at our announcement last week on Trainum three, which is a third generation of trainum that is 40% more price performance than Trainum two. So that kind of uh uh uh step function improvement in price performance is where I think sustainability will be key at the same time. On in parallel in our data centers, we are trying to use as much renewable energy as possible. We have partnership with a company where we are absorbing the carbon that's getting emitted. All of that is key for making computing more sustainable. So and I think as a whole, the industry needs to look at it even more carefully and its model optimization because some people worry that OK, you make the models more efficient, then people just use more of it. It's called the GS paradox, you know, and actually the overall consumption is higher. Is that a concern? Are you seeing that also as you drive down model price, are you seeing usage actually increase over this actually validates what we did with our own models? If you look at the model family, we launched with Amazon Nova, we are taking a lot doing a lot of work to the capabilities of the large model into the smaller font factors or smaller sizes to a suite of techniques. And this is where if you look at where you're going Jeremy with it, like inference is becoming a building block that every application in the world will use. This is why if you thought about it as a per request inference or a per token request, each of that has to go down by a factor of factors. Right. I mean, and of late, we have all we managed to get an order of magnitude faster with our influence and more efficient, but we need to continue doing that work and I'm very optimistic that we'll do so because our customers want it. Ok. Right. This is where instance as a building block is such a key element for us to think about as you're building applications. Thanks. There's a sort of war for talent going on out there. How is Amazon sort of competing in that war for talent? How are you able to kind of retain the engineering talent you need? I think one thing I cherish about Amazon that it has uh world's most customer obsessed scientist and ML talent in the world. Uh because here we are looking at what's the best way to solve and build real practical A I solutions. So I'm very proud of the team that we already have. Uh We just today announced our uh San Francisco A I hub, which is very crucial for this. Like we have had a present in Sunnyvale for a long time, but we said we need to be in the city. So this is why we are in the city now. That's why I was also here today. And so that I think it's great times for A I scientists and A I practitioner. But the good news for all of you in this room is that the bar to build with A I has suddenly reduced. You don't need a phd in machine learning or mathematics to build with A I. This is the way where you now can just prompt these systems or run out a lot of experiments where you don't need to be the A I expert. And frankly, I believe that more and more work will be now at the application layer as well where you will see that again. A phd in machine learning is not required if you could tell the audience a lot of whom are trying to figure out how to get from A I in their own businesses. You know, if there is just one piece of advice, you could give them on how to do that, what would it be? Well, first two things, one, we're gonna run out of time. So first, I think we are seeing a phase where generative A I is moving from experimentation to real production use cases, which is very positive. And what I'm seeing is that if you look at our development environment, we now can save 4500 hours of SD software development engineer time because you don't need to do a Java migration with a lot of people, you can simply do it with the five people in a room. That's incredible. So my last minute advice for you on that one would be focus on getting your data right, your decisions and your daily workflow captured. Because there's the magic in A I is that once you have the data and the decisions, it can learn if you don't have the data and the decisions it's not gonna work. So focus on getting to get your data right. I'm afraid we're out of time, but Rohit, thank you so much for being with us.

Presenter: Bill Gross, Founder and CEO, ProRata AI; Founder and Chairman, Idealab

Brainstorm AI 2024: Building A Better Model For Creativity

I've been an entrepreneur all my life. And as a 16 year old in passing to California, I was reading Fortune Magazine looking up to all the incredible businesses that they spoke about. And then I had a chance to come to all these fortune conferences and learn so much from all of you and all the people at Fortune Magazine. And now here I am on stage telling you about a new idea. So it's really thrilling to be here. Thank you so much. What I want to talk about today is the disruptive world we're living in right now, which is just more disruptive than anything I've ever seen. You. Cue to my slides, please. We are living in wildly disruptive times. And the question is how will creators survive in these times? Think back to all the different technological revolutions and how creators how to adapt to be able to thrive back to the printing press, radio, television and of course the internet. But I think these times are even different than anything we've ever seen before. I think this A I disruption will dwarf anything we've seen in our lifetimes. I put up on this slide four things that I saw in my lifetime which when I saw them I saw thought the world will never be the same. I go back to the IBM PC in 1981. It actually came out the very month I graduated from college. I went down to my local computer land store in Pasadena, California bought a $5000 monochrome machine with two floppy drives and 256 K of ram for $5000 no hard drive because the hard drive hadn't been invented yet. But it completely blew my mind and of course, it changed the world. It led to a whole new form of creativity code and led to trillions of dollars of wealth creation. Then Netscape in 1994 when I saw that, I also felt the world will never be the same. In fact, I was running a company called Knowledge, a venture at the time selling educational software for kids. And we sold that company in 1995. And now I had all these new ideas for internet companies that you could build on top of the browser. I put down my favorite 12 ideas and I was trying to select which one I wanted to start. And I actually decided to start all 12. And the way I did that was by starting idea lab, a technology incubator where I could start all 12 at once under one roof. One of the companies we started based on the browser was a company called Go to. We invented paper click advertising and it completely changed the way monetization worked on the internet. We were really proud of that, but that just shows how new models are needed whenever there's a new platform, a new technology. And of course, the mobile revolution has been talked about already today that changed all of our lives in 1994. When I started looking at Netscape and seeing how powerful it was, there were 30 million browsers in the world, think about how pitiful small that is compared to the number of mobile devices that are on the planet. Now, 5 billion supercomputers in everybody's pockets all over the world. So all these changes were so dramatic. And then November 2022 we're at the twoyear anniversary of CG BT being released. I remember playing it with that, that one day and saying this is going to change everything. And of course it has, it's led to a trillion dollars of wealth creation and expenditures in just two years, not like the PC, which took two decades for that kind of impact to be felt. So when you have something that big, why does it cost so much change? Well, hugely disruptive changes occur when the cost of something approaches zero with the internet, the cost of distribution approach zero with the cloud, the cost of storage approach zero and with generative A I, the cost of knowledge or knowledge creation is approaching zero. So what does that mean for content creators when the cost change that dramatically, this disruptive moment is really, really large. Well, until now one side of the equation, some of the gene A I companies are saying the content must be free. The other side of the equation, the content producers are suing on the one side, you have Mustafa from Microsoft saying everything on the internet can be used for free to train A I models or Sam who is saying we need this copyrighted material for free to make this business model work. On the other side, you have the New York Times, Dow Jones and many, many other companies suing a lot of the generative A I companies because they feel their content is being stolen in a funny tweet Ariel Dumas says I've been told to stop stealing muffins from the bakery. Unfortunately, it's the only way to keep my lucrative muffin stand of business. Everyone is fine with this and of course, there's sarcasm here, but it's maybe not that far from the truth about companies lifting other people's content and using it without paying for it. And that leads to a dilemma today. Well, in almost all other forms of digitization revenue is shared, I'll give you some examples right here on the App Store and on Google Play revenue shared with creators on Spotify revenue shared with the artist. In fact, Spotify shows that you can make a business like this work even with significant revenue share, Spotify's market cap just today crossed $100 billion. They had $16 billion of revenue and $10 billion of revenue was shared with creators. It was shared based on counting the streams that were played youtube shares revenue, full 5050 revenue share sharing revenue with the street, with the person who created the stream, Apple music and also Apple News. Apple News is another example of another method of counting. You can count by streams, you can count by views or downloads. Apple News counts by minutes minutes spent in different media content. And Apple News took in $2.3 billion of subscriptions last year and they paid out 1.15 billion to all the content partners who provide the content and even in the physical world for offline distribution revenue is shared companies like Nielsen Ascap BM I count and share revenue in those use cases as well. So why are there no revenue shared standard in generative A I? It's because there's no simple streams or views or downloads to count the answers that come from generated I generative A I are an amalgam of different content. But what if you could reverse engineer the output of generative A I and figure out what content contributed to the answer that would enable a fair revenue share. So right now, some of the A I companies are saying we don't want to pay because we can't afford to. We don't think we have to. But some generative A I companies are saying we can't pay because we don't know how to. Well, what if there are a way to solve how to, well, that's what we set out to do. We set out to create an algorithm that could take a look at the operative gen A I, whether it's text images, movies or music, analyze it using A I and figure out where that content came from. And not only figure out where it came from, figure out the percentage that came from each place. And just today, we launched this new search product called GIST at G dot A I. You can go type in a query and you see an attribution bar at the top that attribution bar at the very top is breaking down the sources and where how they contributed to the answer. So for example, in this particular question, 34% came from the Atlantic, 25% came from time, 25% came from Fortune. Fortune is one of our partners and 16% came from the Guardian and all of the sources are called out clearly on the right. This then allows us to share revenue 50 to 50 with all those partners on that exact pro rata basis. We think this is fair. We think this is a great way to encourage monetization of content in the generative A I era but also encourage by providing a business model for more creators to continue to make great content to power. Generative A I. If generative I only starts using synthetic data for its training, the result in quality will go way down. But if content creators have the incentive to keep on creating great content, the quality will go up. So we really believe this is the future and we really want to try and make something like this be the standard. We really feel that generative A I needs to be ethical and fair to make this grow and thrive. I showed you the example with text. It also could be done for images. Here's an example, the prompt that was typed in was generate an image of a mass red superhero that can climb walls, meta generated this image. We look at that image, we analyze it, we look at all the source images and we figured out that this came 90.3% from Marvel and 6.2% from DC Entertainment. Again, giving us the exact percentages so the revenue can be shared. So engine of A I, it is now fully possible to look at the output and figure out who contributed uniquely to that content. So far, hundreds of publications, authors and musicians and artists have joined our system. Actually 400 as of today, 20 more just joined this morning, we made an announcement this morning of new ones including a company called Lee Enterprises which has 71 markets around the United States. And we're working with them to revolutionize the way local news is surfaced in generative A I and also the way local news can monetize with generative advertising as well. We really feel that this is a new frontier. If you think back to search and how it's gone through multiple iterations, search 1.0 was monetized with banner ads. Think back to excite and Lycos and Altavista and Yahoo search 2.0 was monetized with paper click. I believe that search 3.0 which we're seeing right now will just deliver you answers directly. You won't have to click but it will be monetized with paper use. The paper use is the ability for the content creators to make some revenue every time their content is used. Of course, their content also can be used for training, but that's a one time fee just for training. The model paper use enables the content creators to have a sustainable revenue model that can be ongoing. So pro rata dot I enables pro rata compensation on this paper use basis. And I really feel I'm so passionate about this because I really want to make a world where content creators can thrive where it's fair and ethical for everybody and everybody can participate and wherever the content is created anywhere on earth, it could be surfaced in answers to help you have the best content possible and to make a revenue stream monetti, I'd love to work with any of you here. I'd love to have some of you join our Pro Rata platform. But mostly I would like feedback and ideas on how to make this a more successful, thriving ecosystem around generative A. I feel free to reach out to me directly. I'm Bill at pro rata dot A. I thank you very much. You've been a great audience.

Presenter: Romain Huet, Head, Developer Experience, OpenAI

Brainstorm AI 2024: OpenAI Demo

Thank you. Good afternoon, everyone. Super thrilled to be here today. It's an interesting time in A I, the pace of change is really accelerating and there are more A I applications going from prototype to production every single day. I'm Roma and I lead developer experience at open and you know, a lot of people know us for Chad GPT. But in fact, our first product was the open API the ability for developers and for companies to take our A I models and bring them into their own products and services. And today, we are very fortunate to work with more than 3 million developers on the platform and 92% of Fortune Five Android actually use our products. So it's an amazing time to be all of us together and building in A I, you know, one of the favorite things in my role day to day is to work with amazing founders and companies who are pushing the envelope of these models every day and also showing them the art of the possible with what we, what we put out there in the world. So in that spirit, instead of having any presentation of slides today, let's jump to my laptop and take a look at some live demos. So first off at the remiss not to mention Sora, we just launched Sora this morning, our video generation model and it's now available on sora.com. Uh So what you're looking at here is the home page with features video already created in the last few hours by the community. Uh It's pretty amazing. In fact, we can fire off maybe a first uh video generation here. I'll say a golden retriever dressed as Santa on Christmas Day, you know, sounds pretty uh pretty cool. Um There is an enormous amount of demand at the moment for Sora. So sounds like uh you know, the service is already at capacity. That's OK, you can check it out later. Uh uh You know, and these are some of the beautiful videos that are already creating. Yeah. So now in fact, to show you something more concrete but with the wind to Sora, I'm going to bring here something about multi modality, multimodality is what we refer to as the ability for A I models to not just not just work with text, but also work with speech, with audio, with vision and also with video as you've seen with Sora. So I wanted to bring you here with a demo of how we can combine all of these modalities together to create something very unique. So this is the only part of the demo that is not live. This is describing a prompt as you can read on the screen in great detail about kind of the history of the German automotive automotive industry. And so if I click here, this is the video that came out of Sora. So what's pretty amazing here is that uh it's only about us switching together some themes out of Sora and adding a little bit of a soundtrack on top, but all of the images are untouched. This is straight out of the model. It's pretty magical. So now let's say we want to use such a video as a basis for a presentation. And what if we want to tell a story that this image, these images in this video can tell us. So what I'll do next is I'll click this button extract frames and I'll slice a few frames out of this video with A Z are generated. And now what I'll do next, I'll bring another modality of GPT photo, the ability to understand and extract insights from images or pictures and that's what we call vision. And so what I'll do next here is I'll keep this button and I'll take all of these six frames. I sliced out of the so video and I'll send them to GPT for all with vision to come up with a story. So don't blank, it goes fast. There we go. So every single time this is unique again, this is all happening in real time as we speak. So now out of these few frames, we were able to come up with that story. So this could be, for instance, the narration for a, for a documentary or this could be something that you can adapt to your own use cases and industries. But next, what if we were to bring this to life? What if we were to be able to bring a voice onto that uh that narration? Now, I'm sure some of you are already thinking ahead here. So what I'll do next is I'll click here to create a voice over and we now have amazing voices in our api like preset voices and they are much more expressive. We'll see that in a minute with the, with the next demo. But here, I also wanted to show you a preview of our voice engine model. The voice engine model is the ability to take a short sample of a voice and turn it into a voice you can use in your apps. So this is only available to trusted partners for now. And we are actively testing this West Dem. We obviously take safety very seriously. We wanted to show you a quick preview of what that would look like. So I'll go ahead and here, I'm going to record a few seconds of my voice. I'm extremely excited to be on stage at the Fortune brainstorm A I conference today and all I need is a few more seconds and I think we should be good to go for me to actually narrate this beautiful video that Sora created. All right. So now that this is so I'll go ahead record it right there. I can just send the script that we created before. Now send this little sample to our voice engine and we'll see how all of these can be combined together to create something a little unique here. So in a second, we should see that exact same video in the heart of the German automotive industry. But now narrated, Rily assemble the machine that would do one more modality that you can play with, right? And every bolt secures the legacy of engineering precision and innovation. Next, I could also do translation. So as you can hear from my French accent, I can maybe also narrated in French. So I'll go ahead and click French here, commitment to quality and efficiency setting. And now this is me narrating this video in French. And last but not least another modality that you can now tap into from our multimodal A I model is the ability to bring the speech to text as well. So here for since most of you don't speak French, but I can actually turn on the caps and all of a sudden you have this other modality of speaking text available. So this is just one example of how you can combine these multi modal models like GPT photo to create something very unique that you can adapt to your own businesses and products and services. But next, I want to show you something else about real time and our real time api so, first of all, here, how many of you in the room have tried Chad GPT with the advanced voice mode before? Wow. Quite a few of you. That's awesome. Well, you know, one question that we've heard so many times over the past few months has been, how can I bring these exact same experiences into my own products into my own services, whether it's to re invent customer service at scale or if it's about like a new way to interface with products? Well, here what I'm going to show you is a loop, sneak peek behind the scenes. So this is the open eye playground. So this is where developers and builders start their journey with our models. And so here I can start a session at any time and actually engage to start to have a feel for how the real time api in the speech to speech capabilities of GPT F come to life. So let's try that. Hi there. Hello there. How can I assist you today? Yeah, I'm on stage at the Fortune Prince, Tommy I conference. Could you welcome the attendees for me? Welcome everyone to the fortune brainstorm A I, so I like how you started, but I think I want you to bring more energy, more enthusiasm and don't forget people watching on the live stream. Welcome everyone to the Fortune Brainstorm A I conference at the Fabulous Saint Regis Hotel in San Francisco. To our incredible attendees and viewers online prepare for an inspiring and groundbreaking event as we dive into the future of A I. Um Now, could you whisper with a very low voice? A secret about San Francisco that maybe only the true locals could know like very, very low like subscription. Did you know that underneath San Francisco's bustling streets? There's a hidden network of tunnels. I did not know that these tunnels were used during prohibition to transport liquor. Discreetly, it's a little known secret. So as you can see the ability to interrupt the model at any time, the ability to jump back and generate intonations, all of that is at your disposal to create amazing ways to interface between your brand, your product and your own customers. Maybe one last example here, I have a test for you here. So here is the deal. I'm going to speak to you in three different languages and I'd like you to speak back to me with those same three languages mixed up. Does that sound good? Absolutely. I'm ready for the challenge. Let's hear it. All right. Chow, Pala Italiano, Francais de Christian stage and I can switch back to English at any time. Your turn. Chow Franci and I'm switching to English now. It's a pleasure to mix these languages on stage with you. So this is the kind of power now you have with these multi modal abilities for speech to speech. And here one example, for instance is that if you want to imagine what that could look like briefly, well, this is what you can do. You can now have voice as a modality to interface with new keyboards. What if I were to say, hey, can you show me on the map where we are? Sure here's Paris, a beautiful city with rich history. I'd like to see amazing Christmas decorations when I'm there. Can you show us a few examples? To me, here are some spots in Paris with amazing Christmas decorations. Galerie Lafayette, the Champs Elysees. This is one way you can imagine how we'll be able to re invent computer to human interfaces in the future with multi modality. And of course, this is not the only paradigm shift that we are excited about because we have also launched the one series of models that can now attack very complex task with reasoning. And we strongly believe that reasoning plus multi modality combined will unleash A I agents next year. And so we're extremely excited about these two paradigm shifts. Thank you so much for your time and I can't wait to see what you build. Thank you.

Christopher Young , Executive Vice President, Business Development, Strategy and Ventures, Microsoft Interviewer : Kristin Stoller, Editorial Director, Fortune 

Brainstorm AI 2024: The Shape of AI to Come 

All right, good afternoon everyone. Chris. Good to be here with you. Thanks for having me, of course. So I want to start off by talking about investments. Obviously, Microsoft is a big investor in the A I space, but I want to talk about embodied A I. So some of the recent investments Microsoft has made is robotic start up figure, self driving car, start up wave. Where do you see the biggest potential for embodied A I? And how does that fit into Microsoft's bigger investment strategy? So what you just described are two great examples of autonomous systems. And a couple of years ago we did some work and we said, you know, this was even pre generative A I, we felt that autonomous systems were really going to be some of the future applications that would really change how we lived our lives, how organizations function. But like when we look at that space, we think it's going to be a combination of self driving cars or self driving vehicles to put it more broadly speaking, you know, I think it's going to be less about transporting people around and a lot more about transporting goods and services around ultimately long haul trucking would be an example of that or just think about all the vehicles that operate in small warehouse contexts or even in a shipyard, you know, as you move goods off a ship onto trucks and the like and then on the, on the look, I think humanoid robotics is a, it's going to happen, that will be something, it will be, it will be quite commonplace. I think particularly in categories of work where it's very dangerous for people. Think about offshore oil rigs or you know, Antarctic, you know, oil drilling in Antarctica, this is going to be a lot of different opportunities I think for us to bring new types of embodied technology into those kinds of categories and it's pretty exciting. Yeah, definitely. And speaking of investments, obviously, we just heard open A I um Microsoft, a big investor. So where does that relationship currently stand? Are you partners? Are you competitors? How do you find a coexistence with us? We're great partners and you, and it was even mentioned by Jason a little bit earlier, you know, we invested in open A I a number of years ago, even pre pandemic. Um over time, we've had an amazing partnership if you think about all the great things we've been able to do together as partners. Um If you look back, I mean, it's just been, it's been a great partnership and I would say one of the more unique partnerships you even if you would see in technology, um obviously, we have a lot to that we're doing together. You know, you even see, you know, Mike at, at our ignite event. A couple of weeks ago, we announced a number of new innovations, all of our copilot innovations are built on with open A I technology as part of it. And so we're really excited about the partnership and we continue to do a lot together. Speaking of ignite, there is a lot about um announcements around copilot including copilot actions. Tell us a little bit more about this and what excites you the most in terms of co pilots capabilities going forward. So three big things at at ignite, there were many, we had many announcements, I think 80 announcements, but three big ones, you just mentioned copilot actions really templates or really being able to bring more agentic functions to co pilot. The second one was agents which is ultimately going to be the extension of how people do their work and integrating that into a process flow. And the third is really Azure A I foundry, which is going to enable developers to build a lot of their own copilot like capabilities, build their own agents. So we were really proud of all the announcements, but those are some of the big categories that we announced a couple weeks ago. I want to come to the audience team for questions. But first, I want to go back to what you said about Azure, tell us about what the advancements in that have been and what excites you about it for the future. So first of all, think about Microsoft as a platform company, we have, you know, we we are bringing a diversity of capabilities to help developers build their own tools, their own co pilots, their own agents into, you know, into act. And that's really what Azure A I foundry is all about. And you're seeing a lot of that, I would say if you look back, we were really one of the first companies through github Copilot to bring capabilities for developers to do more code development. Now, what we're doing is bringing capabilities and tools for developers to build agents, to build co plots and to do that using a diversity of models. We have a lot of different models on our catalog and are allowing developers to use what works best for them. Make sense. Any questions for anyone over here? All right, good because I have a lot I can keep going. So um earlier last month, so Microsoft unveiled a new open source multi agent A I system system called Magnetic One. How is that different from all the other multi agent systems out there today? And where do you see the most commercial success there? So what you're referencing is some work that came out of our Microsoft Research Organization and they published a paper on on multi agent, which is an agentic framework. And if you think about it, it's really based on a lot of the theory that, you know, agents will function in teams in many ways, just like people do. And you'll have a lead, for example, a lead agent that will direct some of the specific work that will happen by very specific functional agents. And part of the reason for that is as you go through any completion of task, whether it takes a period of time or whether it has more both facets to it, there will be changes, there will be adjustments that need to be made and the lead agent will direct that work similar to the way a team leader might lead work in a project. And so it mirrors that concept, that concept. But if you think about it this way, ultimately, agents will need to function in concert with one another because there will be many different agents that work in an organization uh that might be even at work on behalf of an individual. And so having a lead agent that directs the functionality of other agents is really the core concept of that. And you know, you've seen a lot of this, you saw a lot of this ignite, we're focusing on, on extending the value of what we're doing with copilots into the world of agents. And, and the, and the multi agent framework is really a part of that. Makes sense. Another announcement this year that I personally found fascinating was the announcement around three Mile Island. Um So Microsoft announced plans to reopen it to um support A I. Can you talk to me a little bit about what the future of the nuclear industry is as it relates to A I? And then also how does that relate to sustainability? I'd love to hear your thoughts. So look, the way we look at nuclear, nuclear is it is a, it is a clean energy source that will become part of our overall portfolio of energy sources that we use for delivering, you know, our data center capacity around the world. And you know, we expect there's going to be a diversity of energy sources, you know, we've made certain commitments as a company to being carbon neutral and even carbon negative in 2030. And so nuclear is a part of that portfolio. And that's, you know, that's really what that announcement was. It's part of a much bigger strategy that we've got as a company. Got you interesting and how are you looking about sustainability as it relates to it as well? So the way we think about sustainability is number one, it's a must. And you know, we expect that we expect that, you know, there's going to be a lot of different. In fact, there's a, there's a recent blog that we publish, which are some of the trends that we see. And one of the big trends that we see are is really, you know, the ability to deliver more and more compute past capacity with a lot better efficiency. So for example, better efficiency around how we build data centers in the first place, better efficiency on how we do cooling in data centers. That's one of the big ways uh in which power gets used in a data center context is that the cooling take, it takes a lot of energy being more water efficient in how we do and how we run our data centers as well. So, you know, we have a series of different methods that we're employing in our data centers. You know, some of them we've been doing for a while, others are new. We continue to work with our partners, all of our partners like NVIDIA A MD, even the in House Silicon that we're building at Microsoft, all of that is with an eye towards making our data centers more efficient, more effective and ultimately part of a more sustainable uh delivery mechanism across what we do around the globe. If you had to pick um three big bets that you would make on A I in 2025 or big trends, what would those be? Well, I'll tell you what we're investing in right now. Um because I as was mentioned earlier, you know, one of my responsibilities is for our venture arm M 12 at Microsoft. And so a couple of key categories that I think are really important. One is broadly speaking A I infrastructure. But if I were to break that down into some core components, one aspect of that is going to be cyber security near and dear to my heart. I used to be in cyber for a long time. But, you know, for example, we've invested in a company called hidden layer and their, their focus is protecting models against being manipulated by third party actors because the models themselves will become a tax surface for malicious actors. And so that's just one example, we're all of seeing need, for example, for data governance and data compliance, data protection. So if you think about this way, as we bring more A I capabilities into organizations, those A I agents, the A I CO pilots are going to want to access information. Um you need better governance, you need identity capabilities. And so we're seeing a resurgence, we've invested in a number of companies reliance A I is one we recently invested in in that space, that's in the data governance category. We also have a number of our own products at Microsoft first party products that are in this category. And then we're even seeing the need for new databases in some cases, like one of our recent investments was a company called Neon. It's a serverless Postgres database. And we're seeing, you know, we're seeing a lot of demand for different database models, different data architectures as it relates to developers, building new applications that are A I first in nature. And so there's so much happening right now, but that broad category of capabilities that are A I infrastructure enablers is a big area that I think is important to focus on. That's where we're making a lot of bets. And then some of the other ones, you just said a moment ago, embodied A I. Although the last one I'd add is I do, I do think there's a lot of opportunity in uh generally speaking in the health care space all the way from patient care and delivery through drug discovery and development. A lot of exciting companies that are using A I. One example I would share is we, we recently invested in a company called Invita uh uh in Beta and in beta. What they do is they map, they're using generative A I to essentially study what happens in nature and look for the natural remedies that nature has for certain ailments like eczema is one that they've got actually a few drugs and trials to, to treat eczema using natural remedies that occur in nature. And so, you know, those are some, that's just one example of many, you know, we've invested in companies that are um using A I as a better mechanism for diagnosing or assisting pathologists and radiologists in the diagnoses of cancer by being, you know, just really good at the screening aspects of this. So I think there's a lot to be done there across the board in health care and it should hopefully raise the bar for health outcome for all of us. And I love that you have this cybersecurity background too. I think it's so fascinating. So I'm going to ask you on the flip side of things then what, what is your biggest fear right now as it relates to cyber security and A I uh I mean, there's, there's near term fears and there's long term fears. I mean, I look, you, you, my wife would tell you that, you know, being married to somebody like me, you have a certain degree of, of paranoia as your baseline. Same with a journalist. So, yeah, but I, I would tell you, uh look, I think there's a real need for responsible A I it's one of the reasons why at Microsoft, we've invested really heavily in really testing and a lot of the, the use cases around, you know, looking for what we think could be model manipulation, misinformation, disinformation. You know, there's a lot of focus on that the last couple of months here and just even in the United States, given the election people were, were on the, I would say really on the lookout for that. And I think we, you know, we came through all things being equal, came through without a lot of too many big alarms on that on that front. So I think we have to be mindful though that everything that we do with models for good can be flipped and turned to be done, you know, be used by malicious actors. And so just making sure we have the right security models, that's why responsible A I is really important. It's one area we've invested heavily at Microsoft and we encourage all of our own customers, our partners to have their own investments and responsible A I for that reason, you know, in the, in the long run, you know, we'll find new things, we'll have new challenges we face. Uh but I actually, I am pretty confident that we're making a lot of the right decisions and we, we'll, we'll work through challenges as they arise. My words. Well, thank you so much, Chris. I appreciate it. Thanks for joining us. Thank you.

Daniela Braga, Founder and CEO, Defined.ai, Benjamin Plummer, Chief Executive Officer, Invisible Technologies , Jonathan Ross, Founder and CEO, Groq, Rana el Kaliouby, Managing Partner, Blue Tulip Ventures; Co-chair, Fortune Brainstorm AI

Brainstorm AI 2024: The Essentials of Building AI

Daniela Braga, Founder and CEO, Defined.ai, Benjamin Plummer, Chief Executive Officer, Invisible Technologies, Jonathan Ross, Founder and CEO, Gro, Rana el Kaliouby, Managing Partner, Blue Tulip Ventures; Co-chair, Fortune Brainstorm AI

Daniela Braga, Defined.ai, Benjamin Plummer, Invisible Technologies, Jonathan Ross, Gro, Rana el Kaliouby, Blue Tulip Ventures

I am so excited for this conversation. Thank you for being here. So A I is making us all more productive, but man, it's so expensive to make. So I thought this in this conversation, we would go behind the scenes and talk about what it actually takes to make A I. So the data, the humans and the compute. And Jonathan, I wanted to start with you. Um People often think about Grok as a hardware company, but you actually insisted on not being on the tip panel at this event. Um You're more like a full stack solution and I more like Invidia actually, and and I wanted to point to one particular data point. So Grok cloud platform is used by more than 650,000 developers to build a wide range of A I applications. So can you explain why this software hardware combo is so important, especially in A I inference, which is what croc focuses on? Well, um most people don't know this but am DS GP us are actually faster than Nvidia's GP U SI actually had a VP from NVIDIA brag to me about this fact and then said it's actually their software that gives them the advantage. So when we started GROK, we actually spent the first six months working on the compiler and only after we were able to take programs and, you know, lower them into something that we thought we could build and run a chip with. Um did we start designing the chip? And so it was a massive advantage, follow on question. So you're often thought of as invidious challenger and there's also a lot of new entrants in the space like cerebra and etched and Amazon. What is it gonna take to win? Ok. So, um you never wanna directly compete. So we like to say that we do inference and they do training, they would love to do inference too, but we do fast inference. And why is that important? Why is fast inference important? Um Does anyone here want slow A I please raise your hand, you know, no hands. So just imagine if you're doing a Google search and it took 8, 1040 seconds to get an answer, right? That would be intolerable. And that's where we're at with A I today. So with us, we're about 2040 times faster than A GP U for these answers. It allows you to do a lot of the system two thinking and, and other stuff, improve the quality of results. Um But speed also results in increased engagement. So roughly every 100 milliseconds of speed up is about an 8% conversion rate increase on desktop and 30% on mobile. So it matters a lot amazing. Danielle, I want to come to you next. So we've known each other for many, many years. It's great that you're here. Um Earlier this year you presented at the United Nations. And I want to share a quote from that um um conversation. The race for data without permission is only broadening cultural language and gender imbalances because internet content is just a reflection of our society full of biases and misinformation. Your company defined A I is the largest uh marketplace of ethically sourced data and models. What is ethically sourced data mean? And why is it so important? Thank you. I'm very proud of that and actually very excited to hear the presentation before on pro pro rata uh A I which is in line of our belief that uh just because data is public, it doesn't mean it's free. This is always what I've been saying. And now it comes to fruition. Uh We uh created an ecosystem of a marketplace of training data that allows everyone to monetize their data. Uh And we as the brokers, in that sense, we vet legally the data, we make sure we can trace it back to the consent of the participant level. We apply a price, we add all the machine learning readiness and we sell it to a willing buyer that cares about brand reputation and not being sued by copyright infringement. Uh But it's beyond that, we have to take care of. There's another area which is the, as you, as you also mentioned, the uh if everything that is in the internet is biased by definition, it's mostly white male generated, uh it's English language generated mostly uh we can never have. Uh it's, it's just always augments the divide of our society. If we don't bring biased data, purposely built uh and ethically sourced with everybody paid on the chain to the models. And finally, the part that is hidden and nobody likes to talk about is is uh the humans in the loop component, which I guess is that about you uh where a lot of this work is being done under what is called digital. Uh Well, digital sweatshops uh in exploitation of people in developing and uh third world countries uh exposing them to very, very low paid jobs and very harsh conditions, especially uh including in the content moderation world, including uh psychological harm. So those three pillars is how we build our world and our marketplace and how we've been definitely want to come back to the digital sweatshop and the humans behind that in a second. But, you know, I'm an investor now in A I companies and you've actually been quite critical of the A I investment landscape. You've um you've talked a lot about how billions and billions of dollars of funding are going in to fund the same people, the same ideas over and over again. And I, I mean, how do we get to this world where we are actually funding underrepresented humans, applications problems? I, I did mention that too in that United Nations talk. Uh, pretty cool. By the way, I highly recommend it, especially, uh, there, if there, of course, this is a capitalist world and venture capital is focused on making, uh, money which, by the way, so far, very few A I companies make money. That's the other ironic part here. Uh The reality is uh if there must be an agreement and this is why it was important to speak it at the United Nations level where investors have tax breaks to, to incentivize them to look into other fields. The, the agriculture field itself has 1% of investment uh in relation to A I and it's what sustains humanity. So everything is very uh unequal in our world. Uh uh I think that's it. It's at the government level, it needs to be uh incentivized with money, which is what people understand. Then let's talk about the human is the loop. So invisible, you're the CEO of Invisible Tech and it's probably one of the best kept secrets, I think in the A I world, uh you're already profitable, you started off as a work for workflow automation company, but you actually also do a lot of work behind the scenes with open A I and you know, cohere, what do you actually do with an open A I give us an example. Yeah. So I categorize what we do in two broad areas, one loosely defined as sort of evaluations and understanding how these models performing, where are they strong, where are they weak, where are the gaps in their capabilities? So that they have a much better understanding of how these models are performing in the real world. And then the second part of that is creating really high quality data that can be used to retrain those models and actually close those gaps. And you can understand that there's a really symbiotic relationship between those two things and the faster you can crank that fly wheel, the faster you can drive model performance and improvement. Now, you also recently and I don't know if this is public or not, but you recently signed on NVIDIA. Um What are you doing with them? And shouldn't you then work with Rock? Yeah. Look, we've um you know, over the last couple of years worked with the majority of the frontier foundational model providers and it's we've really specialized in the sort of leading edge of that frontier, the most complex work. Um the most sort of intricate training of those models. And so any company looking to build a world class foundational model is either a client or a potential client. Um And given this is so uh exploratory in terms of trying different techniques and still very much in the R and D phase. A lot of that is co creating new capabilities. We're working with researchers understanding their goals and, and building evaluation suites or building data that they can use to actually close the gaps in model performance. Um I'll come to audience questions uh soon. So tee up your questions, please. I want to come back to this digital sweatshop because you employ thousands of humans around the world to basically do all this work. How do you think about what Daniella said? Yeah, it's really interesting because for us as a business, we've never really felt conflicted about needing to sort of decide what's best for our business or what's best for the people that participate in that ecosystem because we know that quality is paramount that the highest quality data is absolutely necessary and to get that you need really happy, really engaged people. And so we spend a lot of time curating this community, investing in them, training them, making sure we understand we're paying well above living wages and really investing and building that out. Yes, it's a good thing to do, but it's also good for business in that these engaged people produce higher quality results. I think probably the most surprising was when I discovered that a few of them had gotten tattoos with the invisible logo, which was probably a little too far in the engagement side of things. But um yeah, really a fantastic community that's loyalty. That's next level loyalty. All right, we have a question here. Can you please say your name and your affiliation of investor in? My question is, you know, having been in the semis space for three decades, I cannot recall the last time a semis start up made it. It's a scale business um working with the foundry and the supply chain and so on and so forth. So while I love LP U and NVIDIA is showing the way Jonathan, how do you think about scaling vis A vis NVIDIA ne MD for that matter. But I don't know if you intended to t me up perfectly, but I think you just did. So I appreciate this. So um A I is a scale game and if you can't get the scale, there's no point. Um One of the unique things about what we did was we actually designed our chips to be using 14 nanometer, which is an old technology. It's under utilized. The fab that manufactures our chips is only about 50% used, which means that next year, if you look at that, if you look at our contract manufacturers, we actually have the ability to scale up to over 2 million of our LP US physically, we can manufacture that. So what we've been doing to get the scale is we've been working with partners. So actually, I've been playing around with this token. So uh if some of you look at me on linkedin. Uh you've probably seen this. So this is my 25 million tokens per second target. Everyone at GROK carries one of these that scale. Uh So when you produce to uh output from these models, they produce tokens and a token is about 1.3 tokens is a word. And so um 25 million is about where open A I and Microsoft combined were at the beginning of this year. And so our goal is to get there, we'll get there by the end of next quarter. So that'll make us hyper scale or scale. But I also happen to have this other one which is um 1 billion tokens. I'll even do this 1 billion tokens. And so this is with a ROM code digital and we're working together to do this. If we do that, that'll be more than all of the other cloud providers combined. And we're, we partnered with them, they're covering our cost to deploy and then we, we split the profits of that and it's a very um beneficial arrangement together because we're really the only ones who can get to that scale. Um NVIDIA has a lot of obligations with existing customers and we can actually build as much compute as all of NVIDIA combined. Wow. You know, I, I have to ask a follow up question. You're very passionate about, about democratizing access to A I. Can you talk about how the scale will be a path to doing that. Yeah, because right now if you want to get access to a GP U, you have to wait in line and it's a long line and it's not a very transparent line. You don't know how long you're gonna wait. Um So if we, so right now Rock gives away about four times as much uh many tokens for free every day as GCP does and we're a start up. And so our intention is to get to a point where we make enough money where we can give access away for free to everyone in the world. Just like you go, you know, into any room here, you plug into an outlet and no one's going to be upset, right? They're not going to say you're stealing our electricity now, for those who are using a large amount of tokens they will pay, but that will cover the cost for everyone else. Amazing. Do we have more questions in the audience? Uh I'm not seeing any right now. OK. Let's talk about jobs. Um What do, do you both see kind of an evolution of A I jobs that are created um in A I that that's needed kind of behind the scenes to train all these A I models and create all this ethically sourced data. Yeah, I mean, we've already started to see shifts and the demand for certain jobs is, is gone up and being created and the demand for other types of job translation and some of these things is, is clearly gone down. And that's not that different from, I think a lot of sort of technology shifts. Um what is different about this one is a, the speed I think these shifts have sort of happened um much more quickly than typically. And this, the second is this really unique attribute about A I and the technologies and their ability to democratize knowledge and skills and capabilities. And so you have people who might not have any technical knowledge can create a website and create these technologies. And so I think it would be really interesting to give these powerful tools to people and free them up to go back to inventing and creating new things and, and removing them from the sort of boring mundane work that holds most people back. I was gonna add two more A I life cycle, more constrained and not uh more philosophically what A I is gonna change. There's uh I like the Xa I calls them A I Twitter, which is a glorified way but really needed way of, of uh calling the humans in the loop uh in the fine tuning process of the model. Uh And uh and the legal, the lawyers, the amount of legal uh with A I, it's a, it's a huge, every company now has to either hire them on in house or, or, or, or contract the consultants. It's really huge. OK. One word answers what is it gonna take for A I to be faster, smarter, impactful and equitable? Just one word that comes to mind. If you have the answer, just say it. Humans, humans. OK. Commitment, commitment, love it. Um I, I think more, more e equitable investment, equitable investment. There we go. Thank you so much to our panelists. Great conversation.

Presenter: Elias Torres, Founder and CEO, Agency

Brainstorm AI 2024: What it Takes to Build A Startup in the AI Era

This is, this is my first time speaking here and it's the first time I got makeup done. So I hope you like it. My slides are not working but OK. Who needs slides? There you go one more. There you go. So I guess this slide is here to tell you that I'm an entrepreneur. I have a problem in my head because I can't stop doing this right. This is my fifth time correction to run it. Uh not five exits. But here I am definitely addicted to this. But in reality, my last company Drift, I sold to Vista and I could have retired. But thanks to the open A I people, the team, they build this great, amazing technology that made me rethink everything and come back because what really happened to me at Drift was that as a founder, I always get to interact with customers directly and I'm just passionate about that and I get to work with, with the customers, they text me, they have direct access to me to my everything and I love doing that right? And that's why I think I've had success because I love my customers. But what happens is that as I get, have success and I grow and I have more customers, I don't have the time anymore for them. And so I think this is kind of probably what happens to you. Tell me how many of you are able to dedicate as much time as you want to, all of your customers raise your hand. Can you do that? What we got to be a little bit more ambitious than that? And so what I would say is that I really struggled with that when I, when I sold the company performable to hubspot, I um I had all these like about 100 customers, about a million in revenue 2009, 2011. And then Hubspot had 5000. And I remember just like when they would call me, I would say sorry, but I have 5, 5000 other customers. I have to build what they need. And so that kind of broke my heart and, and that same thing happened at Drift, right? And so what I'm doing now, right? It really um thinking how I can solve this problem. How can I help all of you take care of your customers? Like if they're the only one anybody interested in that? Right? That's it. Nobody. OK. Just but you gotta like say it, I'm Latino here like we gotta like show your emotions. And so what I'm really here to tell you, right? I think that the title of my talk is like, how do we build companies in the age of a I? So I just want to leave you with some basic thoughts uh in the short amount of time that I have, right? That is like the first thing is that, are you ready for this? Forget everything that you know, question everything that you hear. I hear a lot of founders right now. You know, I'm fortunate, very fortunate for an immigrant come here when I was 17, long time ago to be able to be in the States here with you today. But a lot of people that are starting companies now don't know. Right? When do you hire your first salesperson? When do you hire your first support person? Customer success? We are always looking for advice. Can you believe this is 2024? We have cha GP T cha GP T pro $200 a month and we still don't know how to run our businesses. It's crazy how hard it is right to know how to scale with the amount of customers and the time that they demand from us. And so I feel like we're always looking for advice and we contact other people and we say, tell me how can I do this, how can I do this better? But I'm here to tell you that we kind of have to maybe if we want to look at the past, we're gonna have to forget look at the past and hear what they have to say but not do it that way because A I has changed everything. I, I am so excited of building a company. I went to Sequoia and I opened my big mouth and I said, I'm gonna never hire more than 100 people. And I'm going to get to a billion in revenue and they believed it. Three B I hope they're not watching this. This is being recorded. And so what I said was something that I fundamentally believe, right? My last company, I had 800 employees and it was really hard to build a sustainable business that way, right? A lot of our companies are not profitable. And so we have to figure out how to do this more efficiently and how to do a better job of relating and connecting with our customers. And so the things that I'm doing in this company, everything is being done from the ground up through first principles. I'm challenging everything that normally entrepreneurs would do when they're building a company. So for example, in the past, I've hired as, you know, besides the founders, I've hired like an NBA person, anybody of you hired an NBA person to do great things, all this stuff you don't want to do, right? It's great. They like, they make slides and they like, I don't know, they make slides and, and they make slides, right? And so like, it's great, you don't have to, I don't have to make slides. But then another company I did, my first hire was a recruiter because that's the best way you need people to build great things. Guess what I hired for the first time this time besides engineers. Anybody guess customer support. No, I hire a lawyer. So now my first employee was a lawyer. So this lawyer, it was too early. You would think you don't need a lawyer when you're like 10 employees. But this lawyer does all the contracts. Her name is Z and she also does payroll. She also does it, she does security because we use a platform called rippling. And so we have to do more with less, we have to hire less and people have to be willing to do more because we have the power of A I, we are not, you know, I I'm doing all the sales. I'm, I'm an SDR, I'm a, I'm an A E, I'm an onboarding specialist. I'm a customer success manager. In fact, if you check me out on linkedin, my title is customer success manager and um and I'm support. So my company, I'm doing things differently. I'm challenging the norm. For example, my company, there is no support at agency that inc there will never be the only email you will use is my email. Every single customer from now on forever will email me because that's my responsibility. I'm not going to let it happen what happened before where I lost, I lost track of my customers. Right. And I'm going to use agency my own company to do so. So we have to spend less and we have to learn more. We have to challenge the norms. You see companies like Klarna that they're ditching salesforce. I didn't say that I didn't say anything bad about be off, but this is public information and they're ditching work day and they're ditching this and they're ditching that and that might not work for everyone but it might work for you. We are living in the future and you have to question those things that you thought cannot be changed in order to build the companies that would generate a billion in revenue. Just a small group of powerful and efficient people. For example, one more thing, it's like I will never use the CRM. Why, why would I buy a fiat? Right? Anybody here uses the CRM, raise your hand if you use the CRM, do you guys like your CRM? Keep your hands up if you like your CRM? Ok. Nobody does. We need to start demanding more out of software. I am here guilty to apologize. I've built software for about 30 years. Uh The makeup is hiding some of that stuff, but the reality is that most of the software that we built in the past sucked. It's terrible. I think we're worried about A I enslaving us. The old sufferer was slaving us. I mean, like if you, you guys got to go home and update your records, put in tasks and reminders to contact Elias to see if he can help you. God forbid. That's terrible. Right? And so I'm not building that, I'm not using that. I'm not building that again. I'm just telling you that we need to demand more of the software builders of the future to build software that serves you and you don't serve it, right? That understands and makes sense of all that unstructured data. You saw Romaine do that. Amazing demo, multimodality, talking to it translate in every language. I have a customer in France, they have customers in Italy, they have customers in Germany and everything is working well because we have powerful technologies like today. So I hope I left you with some thinking to do to challenge the norms, think differently and demand more from the tools that you use today. Thank you very much.

Chandra Donelson, Chief Data and AI Officer, United States Space Force Arvind Srinivasan, Chief Technology Officer, Maxar Moderator: Diane Brady, Executive Editorial Director, Fortune Live Media and Fortune CEO Initiative, Fortune

Brainstorm AI 2024: A Galaxy Of Uses For AI In Space

Chandra Donelson, Chief Data and AI Officer, United States Space Force, Arvind Srinivasan, Chief Technology Officer, Maxar Technologies, Diane Brady, Executive Editorial Director, Fortune Live Media and Fortune CEO Initiative, Fortune

Chandra Donelson, Chief Data and AI Officer, United States Space Force, Arvind Srinivasan, Chief Technology Officer, Maxar Technologies, Diane Brady, Executive Editorial Director, Fortune

Hello. Uh Good afternoon Earthlings. I thought we should say we're going to go where no algorithm has gone before. I have two incredibly cool speakers here. And I wanted to start actually, Sandra. I want to start with um what's top of mind for me? Which is what keeps you up at night being in space? Go Diane. That is a really good question. So, and I wanna pivot this over to the, the audience. OK. Are you guys, you guys wanna interact with me today? OK. That's what I'm talking about. How many of you guys use GPS systems and navigation to get to and from where you wanna go on a daily basis. I need some energy who make them jump up? And I say that yes, how many of you depend on weather data to be able to make your plans for the week? I'm not feeling scared yet. By the way, Sandra, how many of you want the Department of Defense to be able to monitor our adversaries activities on the ground to be able to protect and defend this nation and your own activities? Oh, I'm just kidding. Never mind. Absolutely. That is space force. Um And what keeps me up at night is knowing that our adversaries, they understand our reliance on space. They understand your reliance and your family's reliance on space based capabilities. And they are taking deliberate actions to develop capabilities to grade our advantage in space. So what keeps me up at night is knowing that I cannot in the United States Space Force. We cannot let you down because you depend on our capabilities each and every day. And now I'm betting that's not what keeps you up at night. Arvin, what keeps you up at night besides the rocks coming at our spacecraft at 27,000 miles an hour there and the solar storms, there's uh uh you also the immense capabilities that generative A I has. Uh, and the ability to create uh concepts from very little information that are believable and real. Uh, Max R is in the business of providing actionable intelligence by looking at the world from space. Uh the ability to, for adversaries to pollute that just a few pixels uh, keeps us up at night. How do we prevent these attacks from, um, from suddenly altering the information that you all live by and drive is, is critical to our information infrastructure from space? Yeah, exactly. So the security, well, I feel like before we get into some of the overhyped under appreciate, which is also interesting. I want to drill down a little bit on what your jobs are you mentioned a little bit. So Chandra, you're the chief data and A I officer. Very cool and Space Force is really only five years old at this point. So what, what do you do in that context come to work every day? What is your role? So my role is very simple um but also very complex and this to ensure that there is not a day without space data. So the United States Space Force last year, we published our first data A I strategic action plan for Fy 24. And it outlined specific actions and deliverables that we were gonna do in partnership with industry um to deliver on our mission um for your awareness if you have not seen that plan and we're getting ready to release our fy 25 plan. Um We have four key line of efforts and we have a very find acronym that it's easy to remember. That acronym is Mars Ma Rs Mars. Absolutely. The first line of effort is how do we mature, mature our um enterprise data A I governance. You guys are all data A I professionals, you understand the importance of governance. Uh The second one is a, how do we advance a data driven and A I enabled culture when appropriate, understanding the importance of being data driven and all of the decisions that we make and then leveraging A I when appropriate. Um The third one is, are, how do we re optimize our data analytics and artificial intelligence capabilities to be able to deliver on our mission. And again, make sure you guys continue to have access to capabilities. And our last one, which I think I think is the most important is how do we strengthen our partnership um with the most data and A I capable organizations to the on our mission. And many of those organizations are sitting here and I'm so excited to be able to talk and engage with each of you during the conference. So um Arvin, I part of when I read about uh max, are you ground proof in real time? Of course, you've got the automated, you know, map making. What's your moat, what do you do? That's really um tell us a bit more about what differentiates you in space, which is, of course, a vast frontier. I'll stop with all of these different analogies. Go ahead. Yeah. What differentiates Macar is that we've invested in infrastructure end to end uh to deliver intelligence by looking at the earth from space to our mission critical partners like the US government and its allies with reliability, resiliency and the ability to survive uh events, both natural disasters as well as uh human made events with uh with the level of reliability that ensures that the information they are getting is accurate and fine. Um So we are going to get to questions from the audience. I I want to just do a bit of a lightning round on overhyped, underappreciated. You can say Elon Musk in both categories if you prefer that. But let me go with you first, Chandra, what would you put as um overhyped over? We are at a A I conference. Um I would say what is over hyped? Is that artificial intelligence is the solution? I'm sorry to all things. I think we need to kind of change the conversation and ask ourselves what is the problem to be solved? Start with the problem, work backwards. And if artificial intelligence happens to be the capability solution and not do that, if not, let's have real conversations about each other about alternate pathways and a solution. OK. How about yourself? For the overhyped? And I do think talking about the role of A I, of course, in this environment is something we're very curious about. Absolutely. Um space, you know, you've all heard space is hard. Um Unfortunately, it's also uh while being hard, it's also a playground for the computer scientists and for the for just any scientist in all of us, what's not to love about it, right? You have this autonomous robot that's flying around taking pictures, you have every opportunity to do cool things with those pictures and use A I. Um what has uh been overhyped is how uh is the perception that it's easy to make money in space? It's easy to turn it into a profitable business. Uh At the same time, the ability for space to transform your world is also underappreciated. If you ever looked at a moving map, either in your car or uh or any other hand held device, chances are it came from an image that max or, or one of its partners? So it's hard to make money, but there's opportunities to make money. So that's, that's a good message to have. Um So I, we have Mike Runners. If you have a question, please raise your hand. Um I do want to ask, I'd be remiss not to ask about, let's say Elon Musk, Jeff Bezos, maybe not proactively with the incoming administration, but retroactively, what impact do you think that um he's had in this play? I'm going to go to you first, Arvin. Um The impact has been very positive to the industry. Uh And, and the reason I say that is that a decade ago, space was not accessible today, we can think about starting a business that gives us access to space on a predictable well-defined schedule. You can get on a launch calendar at a cost, that's a fraction of what it was a decade ago. Um And so it's created this ecosystem where space is democratized for uh for the right players who know how to turn it into a profitable business. So that ability didn't exist anything you'd want to add. I would just say for the United States Space Force, we're going to partner with the most innovative companies that can deliver the capabilities we need regardless of who they are. Ok. Good answer. I think the good answer, any questions that anybody wants to ask, I have so many questions but um just raise your hand, we have micro is just the one back there. No. OK. That's good because I have lots of questions and my next one is basically, I love the idea of um in any conference like this, there's the blue sky element. So the next say 12 to 24 months, what tell us a little bit about what's around the corner. Of course, we're talking in the context of A I but technology writ large. What can we look forward to? Can you give us a preview, Chandra I would say from the United States Space Force. Earlier this year, we did publish our commercial space strategy. If you have not seen it, please go take a look at it. But it really gets to what the Space Force wants to partner with industry and academia about uh to get off their specific problems. Um One of the things I would say, kind of a Moonshot uh thing, but I, but I do see coming online is our ability to leverage more commercial satellite imagery. Um that is going to be huge, it has the ability to truly change how we operate um and how we make decisions. So I imagine a world where uh a guardian or airman um can have a specific question about something that is taking place on the ground and being able to log into if it's a portal and ask that question, be able to get the energy back to enable them to make a real time decision making. I also see that time shortening down. So some of these capabilities, it takes, it takes weeks, sometimes we've gotten things down uh hours, but I see us getting down to minutes and seconds uh very soon. And again, we have the brightest minds in the world sitting in this room right now. And I would love to talk to you and your organizations about how you can help us get after that. How about yourself? Give us some of the sense of what's exciting to you in the next year or two. Yeah, this is A A I conference. So it would be remiss if I didn't use the word A I at least once. Uh the uh but we're at Max, we're excited about the potential for A I uh to change how decisions are made from looking at space. Uh What we found in, in the last five years or so. We've been in the A I business for a while. Is that more images are discarded because they were manually looked at by humans and filtered for information and, and our partners were missing critical information. Uh We see the advances in the ability to apply concepts like object detection and automated correlations between events as making that much more actionable in real time for, for our customers. As we talk about the evolution here from um you know, generative A I toward agentic A I, some people call it identic A I as well. Um How does that impact the thinking? I mean, when we have agents in space again, I think about some of the things that may keep me up at night. Um How does that work? Is it, is it, is it fair to talk about it in that context? So I'll say that we leverage artificial intelligence and a variety of space capabilities, but I will say that when it comes to high stake things and, and missions, we will continue to leverage a human in the loop. Um I do think there are a lot of opportunities in the United States based for us to leverage um agents to be able to help with business operation processes. One thing I think about in particular is travel, traveling. We do a lot of traveling. It would be great to be able to leverage um artificial intelligence to go out and book flights and understand because many times in the in the Department of Defense, we have to move very quickly. It's not uncommon to have 72 hours notice to be able go somewhere. So how do we leverage uh artificial intelligence capabilities to help us more in the administrative business out of the house versus mission. And I think for the people down here on Earth Arvin, um I'd love to get your thoughts as to what are some of the advances that um have real world applications, you know, for some of the folks in the audience, some of the businesses because I think we talk about obviously satellites have a huge impact, but anything else that you think up there, it's going to transform how we do business down here? Absolutely. Uh We have some exciting new advances in how uh imagery from space uh is enabling rescue missions in enabling uh the navigation of drones in at low level uh flying uh conditions and in conditions where uh you can't send an aircraft, you can't send humans uh in areas where uh you have either denied services or the inability to actually physically send someone. Uh We have the ability to map that in near real time. Uh And so this uh has had impact in hurricanes and floods and situations that you would think that um that you can just send an aircraft. But if you have an active storm, it's uh it's extremely challenging. So, Chandra, I know we just have a few minutes left but I'd be, you know, you're the Department of Defense, obviously, um with regard to cybersecurity and nation state actors, you get a front row seat and I know of course, there's only so much you can say. But could you give me some sense of the landscape that what's happening right now that you is on your radar, you'd put on ours with regard to what's happening with some of the other nation state actors. Hey, we all get it. Cybersecurity is important. We all also understand that any adversary is going to try to exploit uh various capabilities through cyber security means. So we're trying to, we're pivoting our mindset. How do you uh secure by design? How do you make sure you're baking cyber security in from the beginning and not at the end of the process. And we're under, we're having a lot of conversations with various organizations about how to do that. We also have our uh cyber security officer at the Department of the Air Force Level um that can deep dive deeper into what that looks like uh for the department. So that's OK. So no, no specifics. That's, that's fair, that's fair enough. Um Well, let's let's go to the part where we just sort of give some of the takeaways for the audience, you know, obviously, um we can't cover the entire landscape right now, but from where you sit, give us some thoughts as to what we should be thinking about with regard to space force, what you're doing and what you're seeing. So I would say um space Domain awareness is very important. So when it comes to artificial intelligence integration in space, our top priority is leveraging artificial intelligence for Space Domain awareness you cannot protect what you cannot see. Space is huge. Um You do a quick Google search. It will tell you that space is around 93 billion light gear large. Um That is equivalent to eight octillion football fields. And I say this because I know eight all of hilly. Yes, I say this because I know Colin is getting ready to come on, but that's eight with 27 zeros, which is great if you're an investor, right? But when you think about uh people having to maintain awareness of what activities are taking place on those fields at all times, making sure that our adversaries are not doing anything that they should not be doing, making sure that we continue to maintain the GPS CAP capability, satellite communication Wi Fi, you name it. That's pretty important. Yeah, fair enough, Arvin. Uh So space has been a niche industry and uh I think we're at an exciting moment where uh there's a convergence of both A I robotics and automation that's bringing it to light. So it's a fertile ground. I think uh where I see the industry going is that we need to partner better with uh both the venture community as well as innovators uh because there's not only a lot to be done here, but it is fertile ground for uh both uh commercial uh exploration as well as growth. Um And so that's the part that I'm most optimistic about and would like to leave the audience with that technology is learned by doing right. We have to create a space for people to experiment and make sure people understand and empower our people that it is OK to fail. Right? That's the only way to get better is, is learning by doing well. And, and you know, somebody who's a sci fi fan, I'd be remiss not to at least ask you um whether it's going to mars extraterrestrial life, you know, I think we have to. Ok, I, do you, do you believe that there's extraterrestrial life? And do you think that's something we should be looking at? Let's just, just get it out there as sort of AAA palate cleanser before we move to our next one and you have 30 seconds or less to give us your summation. We are extraterrestrials for someone else out there, so I'll just leave it there. All right. Fair enough. I'm going to end with, with the Space Force, Chandra. What do you think? Hey, what I'm going to say, I'm not going to answer that, but there is a, there is a organization in the dod that handles those particular questions and um that mission area. But what I will say is that I know that there is a TV series, a funny TV series about the United States Space Force. But what I wanna leave you guys all with here today is the mission of the United States Space Force is real every day, thousands of guardians, men and women in uniform are working around the clock to ensure that you and your families and your organizations and your businesses continue to have access to space data and I am so proud um that we have the ability to serve you. Excellent. And I will point out that Andrew Nusa looks like he should be on the set of men in black today. So that helps too. Please join me in. Thanking our terrific guest. Thank you. Thank you Marvin.

Dave Clark, Founder and CEO, Auger Sanjay Radhakrishnan, Senior Vice President and Chief Technology Officer, Sam's Club Moderator: Jason Del Rey, Tech Correspondent, Fortune, Co-chair, Fortune Brainstorm AI

Brainstorm AI 2024: Smart Supply, Smarter Store

Dave Clark, Founder and CEO, Auger Sanjay Radhakrishnan, Senior Vice President and Chief Technology Officer, Sam's Club, Jason Del Rey, Tech Correspondent, Fortune, Co-chair, Fortune Brainstorm AI

I like it. Who was that? Oh Yeah, I like it a lot. When I hear supply chain, I do the same thing every morning, every morning, Dave, it's great to be on stage together. Now that you're out of a zone, you're, you're allowed to do this. I'm totally unburdened by what it's been. Make your, I'm a founder. I could say any kind of crazy shit. I feel like today. I mean, beside it perfect. And I said I, I come in the jacket and uh I have two great looking gentlemen, but in a little more casual attire. So whoever told me to wear a jacket, you're fired. Um Sanjay, great to meet you as well and have you up here. Um I guess we should actually talk about the content, right? So Dave, I'm going to start with you um over 20 years at Amazon most recently as we know consumer ceo um you've now raised $100 million and I think you're calling it the series a round for a new start up uh focused on creating an operating system for supply chain software. Uh Can you explain a little bit of what is going wrong in the supply chain space and how A I plays a role in solutions. Well, you know, there's been a tremendous amount of investment in this space for decades and especially through COVID, there was a tremendous amount of money that went into the space. But, and so you have a lot of interesting niche players that are there. Right. So most everybody runs an, er, P software and then they run a half dozen or a dozen other software that does demand planning and transportation management, what have you but they don't talk to each other. And so you're left in this world where everything you know, is sort of output an answer and they have to post process this in some form of Excel or Tableau or smart sheets or something uh with the teams of analysts to try to come up with an execution plan. And our intent is to basically go in and allow them to be able to connect all these systems in an integrated way. So you can actually operate the Indian supply chain uh in a longitudinal continuum as opposed to what they're doing today. And supply chains are always global supply chains, at least always difficult with uh circumstances arising out of your control, right? And sometimes uh political related uh other times, maybe a global health crisis. Now, you know, we have with the Trump administration coming in some questions of what tariffs might look like. Are there solutions? Out. Are you hoping that part of the solution that you're building is something where you take a little bit of the uncertainty or at least planning uh in a more succinct way? Well, I can tell you like, e every, every supply most, every supply chain professional around right now is trying to answer the question for somebody, their board or the CEO or somebody said, hey, if tariffs go to X, what, what happens to our supply chain? Like do we need to re do we need to change the who our supplier is? What does our, our flow need to look like? What does their inventory look like? What are the ramifications of, you know, country X putting another an alternative rule in place that blocks something else and you know, everybody's, you know, everybody's using chat G BT or something and getting an answer really quick and they wanna know why the hell their supply chain leader can't come back with this answer. Meanwhile, there's 50 people grinding out spreadsheets trying to come up with these answers. A and what we wanna do is try to is we wanna get all that data in one place so that you could actually ask those questions. And this is where the, you know, the larger models I think will really help us is getting the full data set together, which frankly most companies, you know, maybe Sam's Club excluded and Walmart, you know, outside these big companies with large technical infrastructure really don't have the ability to, to get their data in one place to ask those questions and to run scenarios with lots of alternatives in them, to get a better view of what might be in certain outcomes or what your decision choices could be given a set of potential out. So as part of your whole wish that you're, I, I know you're building the technology right now, but this would be a perfect time to launch. Well, the, the one thing about supply chain is that there's always something going wrong somewhere. So I'm not particularly worried about it. Like uh I've been a supply chain for 20 some odd years and every day is just another cluster of some kind. You can finish that term if you want. OK, you trying to be friendly. But the it is a, every day is some form of mess. It's just what is the mess and where is the mess and who created it? And how do you figure out how to get out of it? And uh you know, I, I, I've been in this a long time and know the pain that operators face every day. And we just, we want to try to find a way to make that uh a, a simpler, simpler uh set of choices for people. So they can really focus on the strategic outcomes uh as opposed to just answering, answering questions for the boss and Sandra. Um I know you have some supply chain background, but I wanted to shift gears a little bit to your role at Sam's. And I've always been fascinated with Sam's Club. I don't know how many people know, but Sam's Club is often the breeding ground for technology progress and innovations inside the Greater Walmart Company. And then often technologies go from testing in Sam's and then scaling to Walmart proper as I call it or the Super Centers Network, we've had a ton of talk about gen A I capabilities, but Sam's club is using different types of A I both in, in helping consumers sort of re reduce some friction in the stores. And also when your products do get through the supply chain challenges and get to your stores, making sure they stay there, they stay in stock. Can you just tell the audience a little bit about what innovations are working for you right now as in regards to A I powered solutions? Yeah, it's great to be here with all of you. Um What I'll say is that we as a company, we are a people led tech powered omni channel retailer. So at Sam's Club, we have over 600 clubs spread across the US and Puerto Rico and we have learned that to win over time, we really need to listen to our members and figure out how to use technology to anticipate and meet their needs. So, um we've been doing this for a long time, uh listening to our member feedback and figuring out how can we use that to make your shopping experience more frictionless in the, in the clubs? And what's number, what's number one on that list? Yes. So number one on that list, few years ago, we got feedback from our members that, you know, waiting in the checkout lanes with the cash registers with a huge pain point. So we delivered this mobile capability called Scan and Go. It's available in the Sam's Club mobile app. Our members can actually walk into the club, open the Sam's Club mobile app, just start adding items to the cart slide to pay and head to the door. Now, when they hit the door, they do have a friction point because you have an associate at the door waiting to check the purchase receipt. So we got feedback from our members that that's additional friction that they would like us to help solve. And so we have come up with this technology called Sas Exit. And it's essentially a combination of computer vision and artificial intelligence that kind of works seamlessly to validate our member purchases and then allows them to exit the club friction free. So for people who haven't seen it, can you just describe there's, is it an arch with cameras overhead or? Yeah. So if you, as you head to the exit of a Sam's club, you'll see an archway and the archway has cameras in it as you walk through that archway, the cameras are taking images of the member's car. Then our CD models kick in to identify items in the cart which is done through feature extraction. So we are looking for things like size, shape, color logos, brands and cross referencing it with our inventory catalog. And once we identify the items in the cart, then we are looking for the receipts from our point of sale system and trying to match the items on the cart to a receipt once we have that match, and this is all in how, in how many seconds it's all happening near real time, real time, right? So you can think about a club where you have the archway is right before the door. So you have a few seconds before the uh the member can exit the club. So in a matter of a few seconds, we do that real time validation and send a signal to the associates standing at the, at the door that, hey, everything looks good and the member walks friction free. This technology is actually rolled out across all of our 600 clubs. And we are seeing that our members can exit at a rate of 75% completely friction free. And it's a huge member satisfaction booster for us. Our exit experience NPS scores are close to a 90. And what it is also doing is it actually giving time back to our associates in the front of the store so that they can use that time and spend, spend it with the members and make sure that the shopping experience is not like I had one question about that because I heard the 75% number and that's more than half, that's a big number. I can do math up here too. And at the same time, that means it's not 100. Right. And so there's a staff member waiting. So I'm curious like when you think of productivity is that one, is that person doing other things when they're not, you know, checking out or they I'm assuming they're still standing by. So I'm just curious how you think about the labor dynamic there? Yeah, I mean, really the role of that uh associate standing at the end of the club, it's transforming to actually greeting the member, thanking the member for your service and, and telling the member that we appreciate the business, right? So the tech that we are deploying actually helps us in that journey because where previously they used to check every receipt now they are actually spending, you know, 75% of time back and they are using that time to actually greet the member and show them that we trust them and we are not judging them and we are creating a shopping environment that is actually more pleasant and trustworthy. I'm gonna come to the audience for questions. I don't know if we have any right now. Otherwise I'll come back. Otherwise I'm gonna, I have no shame in calling on folks that I know in the room. So, um maybe right here. Uh So I on the labor point and, and Dave, I'm curious for your take on this too. Um Obviously, when I, when I hear these discussions or taking part in these discussions, productivity, that word, I even used it. I'm surprised myself. Productivity comes up and automation comes up and there's a natural labor question, Dave, I'm I'm curious with the development of the operating system, you're talking about like how much as you start to talk to clients is labor something they want to solve for. Um And just how do you think about the impact of what you're trying to build on workforces of your, your end clients? And I think you get to look at companies that are always trying to take cost out of the system. Are you trying to improve service levels at lower costs and the big cost drivers for most companies? I mean, labor is a big chunk of that. But when you look at cogs transport, like there's a lot of other costs in there that tend to be bigger frankly than a lot of the the unitized labor costs. Um particularly when you start talking about things you can automate out. And so from for what we're building, it's much more about reducing the overall cogs in the system. So taking working capital out by improving terms, you know, getting cycle times faster, pulling inventory waste out of the system is a much greater sort of driver of bottom line impact to most organizations unless you, you know, unless you're really talking to wholesale fundamental uh sort of labor redeployments in some form. So for us, it's much more about working capital, lead time adjustments, transportation optimization. We think there's a lot, I think there's a lot more interesting dollars and in that space and right now than there is in labor, the robotics is gonna drive a ton of labor uh savings in that over time. But, you know, our focus is really about the much more of the working capital utilization and transport. And is that also what you're hearing from, you know, fortune 2000 companies or they, or, or maybe in public, they say that. But, you know, when you're in a room just with the chief supply chain of they want both. Like, I mean, everybody wants everything right, like, you know, everybody wants to give customers everything and extract the maximum cost out of the system. And so, you know, there's always this, you know, unfortunately, we've made productivity a bit of a dirty word and, and, you know, we sort of make these, you know, taking, you know, we've had productivity things in place for 100 you know, for a long time, we have humans and we keep finding new and better ways to engage the, you know, people and people's brains and what people do for a living. And I think we're gonna continue to do that. So I'm personally not, I think we, you know, there's too much hand wringing over that and worry about it. Uh But I think when I look at supply chains, the labor savings is interesting but the total end to end waste and sort of overproduction under utilization and transport. Um you know, lack of forecasting capabilities, all those things that both planetary waste, human waste and cost dramatically outweighs uh what the opportunities are and sort of front line labor safe. Do you have any questions in the room? I'll keep going. Nothing yet. I'm going to call on someone. I will. Uh Sandra I another technology I know you all are, are have been talking about. That's A I powered is um you have what, what I think used to be called scrubbers, basically um mobile robots that Queen Queen, the the clubs and attach them. Now, I think is it in some stores or all stores? Um Our cameras also and these are helping with sort of out of stock issues and just giving more knowledge to employees on inventory. Is that, do I have that Moly? Correct? Yeah, you have it right. So um uh Josh inventory management in a physical retail environment is extremely complex, right? Because things move around all the time. So a few years back, um we fitted our autonomous scrubbers with cameras and these scrubbers traverse the aisles of our clubs, taking pictures of merchandise on the shelf. We take an average of 24 million images a day and we have CV and A I models that actually extract information out of these images and convert them into insights and inventory intelligence that we feed back into systems that our club associates use in their handhelds and we use it to drive the next best action for our club associates. So for example, if our members Mark Peanuts is running low on the shelves, we can actually generate a signal to our associates to drive a restocking action. Or if the price sign is incorrect or if price sign has fallen down, we can drive actions to our associates to go fix it. We also use it to actually update locations of items so that the next time the associate wants to find that item, it's, it's much easier and we can send them to the right location. We have also added RFID readers to the scrubbers. So as they traverse the aisles, they are reading on average 180 million tag reads a day. And we can drive actions like if the small size of a T shirt is running low on the apparel rack. But we find that inventory is sitting in a box below the rack, we can drive a restocking action and ask the associates to go restock the small size T shirt back on the shelf, right. So using these solutions, these inventory management solutions with A I models and some other innovations, we have eliminated in excess of 100 million tasks from our associates lives in the clubs. And that's time back to our associates to spend with our members on the floor. I just want to make sure I heard you correct. 100 million, 100 million tasks. I'm I'm trying to, trying to think of how, how a role could have that many times or this across different roles in the club, across different roles in the club. Got it. Um One more, one more call for questions. No, you're my man. Can you tell us who you are and where, where are you from? Oh yeah. Right. Oh Yeah, right. Here in the front, we have a microphone, we can hand this gentleman, the mic runners gave up. Uh Yeah, yeah, we can. Do you, do you mind standing up? OK. What was the friction from your associates when you were employing this technology? Because there's always friction as new technology gets deployed, especially for lower skill workers. Yeah. What I'll say is um we've done this enough that we know how to roll and roll out and manage through the change management. So we have different um environments uh at the home office kind of lab like environments where they can simulate activities like this. And then we have smaller size clubs where there is like there are members coming in and shopping, but it's a much smaller scale club where we can actually start off and run the first pilot. Uh We also have the option to actually run some pilots on a couple of aisles. And so we have a mechanism to kind of kind of slowly bleed this in, test it out with our home office associates first, then take it to one club, a small size club, then take it to a larger club and then test it and then roll it out to scale. So change management is, is an issue. But we have done this for a while that we have uh uh you know, figured it out. Um on that, on that topic, I'm always curious how you go about and Dave maybe from Flexport or Amazon experience might have some thoughts as well. How you go about getting real transparent, sorry, real candid answers from employees about how new technologies are impacting them for better or worse. And there might be more, more willing to give the open uh the positive answers, right? But if there's a manager or several layers of management between them and whoever's receiving the answers. Um uh My, my experience with the opposite of that, which is usually you can get the negative feedback really quickly. The positive feedback started again, completely agree. Feedback is, I mean, our club associates know how to do their job. So when we uh give them tech that works. We hear about it immediately when we give them tech that does not work, we hear about it really fast too. So they are very open with feedback. OK. Change is bad, even good change. And so a a anytime you roll out something, I mean, this is in my experience, no matter how phenomenal you are rolling it out, it is always bad for a little bit. And so you get the negative very quickly I find and, and, you know, the hard part is I found is working through, keep pushing through it to get to what, you know, what is fundamentally broken versus what, you know, what's just learning pains and, and being able to sort the difference and the one path is to execute it over, you know, sort of do the matsa, you know, step it up over time now, something bigger the other is burn the boats and throw it out there and everybody fig, you know, you only have one choice but to figure it out. Well, next time I live in the ladder in the ladder, burn the boats. I think next time, next conversation we're gonna have us on burning boats. Um, we're out of time now. Thank you, Dave. Thank you, Sanjay.

Colin Kaepernick, Founder and CEO, Lumi; Super Bowl Quarterback, San Francisco 49ers Interviewer: Andrew Nusca, Fortune

Brainstorm AI 2024: Colin Kaepernick’s Next Revolution 

Hey Collin, how's it going? It's great. How are you? Good, good. We uh we got a lot to talk about my friend. Um So why don't you explain the origins of loom me? Because I was fascinated to find out that it's really focused around comic book, illustrators, graphic novels, things like that. And I thought Colin Kaepernick, so please indulge us. Yeah. So part of this came out of 10 years building a media company, five years building a publishing company um and seeing the challenges that creators face. But the other part was I grew up on comics. Um And we had a whole stash of comic books. We had boxes and boxes of them. Um My brother is actually the one who really got me into them. He's 10 years older than me. Um We used to have a sounds a little sketchy now, but there was a house down the couple blocks away, we ride our bikes to and the living room was just comic books. Um So we go through and grab the ones we wanted and go back home. Um But as I went deeper down creating graphic novels, being on the publishing side and seeing the challenges that creator are facing, whether it's writers, illustrators, anchor, um colorists. Uh What we found is a lot of them are living paycheck to paycheck project to project. They actually aren't creating things that they want to. Um Because they're just looking for that next gig and a lot of them have their own ideas that they want to bring to life and are just looking for the avenue and opportunity to do so. Is your company focused around the kind of creative tool and support part, the distribution part? Because that's important with publishing too, where does it come into play if you're one of those folks? Yeah, both of those. OK. Um So very focused on creating that tool set for creators to give you A I assistance where and if you need it. Um So you can go on the platform and create and never touch A I if you don't want to or you can use it through the process to help you. Um And then that is distributed directly on the platform. Um trying to solve that distribution issue that a lot of creators face. How does this company make money? Because that's the number one question when you're talking about media and publishing? Absolutely. Um So a few different ways, the first way is we do a monthly subscription. Um that's our immediate revenue generator. But as we look more broadly, we have on-demand publishing for physical publishing publishing. We've realized creators want that they need it. Um And it's a way for us to do that profitably from the first copy sold, which is a anybody that has worked in the publishing industry. Um you know, it has a very difficult space to work in and be profitable. So we solve that. And then the last piece is we help with the merchandizing of the stories that they're creating. And I've released graphic novel, multiple children's books. Um I talked to a bunch of artists around this and what we've realized a lot of them aren't merchandizing very well. And I actually didn't merchandise my own uh projects because it wasn't an easy process to be able to do that. So that's something that we're looking to solve for them as well, right? What's harder publishing for football? Depends which part. I mean, the physical tool, football definitely harder. Um figuring out how to be profitable and scale a publishing company is no small fee. Yeah. Fair, fair, fair, fair. So what is the status of Lumi right now? Where are you at in terms of a product? Where are you headed? Yeah. So right now we have a product beta testers on the platform. Um Interesting enough, we actually have contracts with school signed right now. Um And they're looking to use that as something to advance learning um project based learning to help literacy, literacy rates, um student voice, student engagement. Um So seeing a lot of traction there in use cases that they want to elevate students with. How are your uh how is your company handling? Ip we've talked about uh copyright and, and concerns about that on this stage already today. Um I think that's the big question throughout the industry, how we look at is creators own everything that they create on the platform, anything that they bring to the platform they own. We're not touching that we don't wanna touch it. We wanna be a space where creators are owning the content because that gives them longevity. And if we're serving them in the best way possible, that creates a much more reciprocal relationship. Um And it also pushes us to be the best company that we could beat for creators. Um We don't want anybody controlled or constrained by that. So that's a great point around that. If someone creates a breakout comic or graphic novel, they could take that and sell it to Netflix or Disney or any other platforms, nothing that we're doing on our end to restrict that in any way. Got it. I'm going to come to the audience for questions in just a minute. But I want to ask you one more. You've already raised funds for this company uh that took you on a, you know, the, the show uh with investors. What did you learn from that experience? Um Yeah, very similar experience to playing football. Um Oh, please go on. Yeah, so what you realize is you aren't gonna be everybody's cup of tea. Um, in football when I was getting drafted, even when I was going to college, you always looked at, it only takes one, you need one person to believe in you and then that gets the ball rolling. Um And ultimately, you want to have a chance to be in the game. We found some great investors that support us in that way. Um So any other founders or people looking to raise funds? I, you're going to get a lot of nose through that process, keep going and just look for that one person that truly believes in you and is going to support you. Yeah. All right. Let's go to the audience for questions, please. Questions for calling. Oh, yes. Right up here. Well, please wait for the uh the mic your name and who you're with when you get it. Thank you. Thank you. So, uh Lorenza Ple and I work or have worked for a while with the economist. So I think it's great what you're doing, especially uh you know, trying to help the community and to people with low resources, to create content to sell, etcetera. Um So I'm curious to, especially because I've worked for a long time, mobile apps, engagement subscriptions, etcetera. Uh Yeah, I would like to understand what do you think is the best way uh to, you know, grow your platform, but at the same time making sure that it's affordable for people uh with low resources. And uh are you partnering with any other, let's say third party companies that will help you perhaps or get them like free credits? Like, I don't know, like to publish on Amazon books or to, you know, like get some sort of partnerships with, with publications to uh yeah, help uh help those with low income. Thank you. Yeah, that's a great question. Um So that's actually part of the foundation of why we built the platform. Um So right now, subscriptions start 20 to $75 a month. So we take a tiered approach to make sure we're trying to give access to people who may be financially constrained. Um The other part to step back and look at the difference in where we're at on this is if you look at the lowest bar of where creators are typically entering a publishing space today, typically is through a Kickstarter and they're raising 5 to $10,000 to get a project off the ground. Um So we're drastically lowering that barrier. Now, if you look at projects that I've done in the past, typically, you're looking at a six figure investment and you're looking at a two plus year production timeline um before it ever sees the light of day. So we're trying to break down both the expense and the speed of creation for creators to be able to get their ideas out into the world. Right on another question for Colin, please. Yes. Right up here in the front. Hey, Steven Steven Wolf Beta. I'm the CEO of AL Fund company. But before that I started a kids media company focused on diverse kids media called Nandos. And one of the great areas for creators to really monetize their I pe is actually with merchandizing. Do you envision the platform actually expanding into products and giving the ability to actually have storefronts? Um Just because there's so much more to just the content, it's actually all the merchandise. 1 1000% you are, you are dead on with how we are thinking about it. Um No, one of the things that we realize is the people who are consuming the content, the people who are buying the books are also the people that are most likely to buy the merchandise, like it's a natural progression. Um Most people aren't doing it. Um And the ones who are actually, they don't love the process of doing it. They just do it as a means to an end. Um So we want to help simplify that process all the way through for the creator from creation, taking that idea to a finished product, to the publishing, to the merchandizing. So one, they can create the things they actually want to. And two, they don't have to worry about the being the business owner, the operator handling all the logistics. Um Every creator we've talked to they're like, I just want to create, I don't want to do those things, but a lot of them end up doing it because they have to today and those are barriers that we can break down. Yeah. Yeah. Yeah. II, I understand that mindset very well. Let me ask you a question. Colin um, you know, you've gone on this, uh, this journey po you know, post, post football every morning, by the way, you still training, of course, OK. You heard it here. He's still training. I was, I was training before I got on my flight to come out here. Now you're making me feel bad. OK. So, so, II I would love to understand, you know, who you see as a mentor for this part of your career, whether that's in Silicon Valley or somewhere else. That's a great question. Um I think there have been a few different people that have really helped me get to where I'm at today. Um One has been in the VC community, Ben Horowitz has been great person to lean on. It's given me phenomenal insight. Um Williams has been another one who has really helped me my thinking. Um Having the opportunity to sit on the board of medium has been incredibly insightful and very helpful for my own development. Um And the other one just thinking about where I'm at today is Alexis Christina, the 776 team um who is our lead investors. They're incredibly brilliant and also great people. Um so to be able to have them to lean on and ask for guidance. Um And also just pick their brains on things that we're working on. Um how we can scale things more effectively, how we can reach, reach creators more effectively. Um Especially when you have someone like Alexis who's built reddit. Um You know, you have a great person to be able to lean on in that regard and an enthusiastic one. All right. Well, let's go for one, one more question. I think there was one over here, but I cannot recall. Oh, it's way over here. Yes. Hey, Carol. Hi, I'm Carol Vanda. I chair um responsible A I for Mars Incorporated. But outside of that, I'm actually an author. So I have a question because I know a lot of authors have negative feelings when it comes to A I, they feel that it's stealing their copyright, their IP how do you overcome those objections to launch Loy? And thanks for your great work. Oh, of course. Um Phenomenal question. And we have talked to a lot of writers about it. Um I think one is understanding that A I isn't going to do the work for you. Um Two is understanding that most of the writers are already using A I, they just don't know it as A I, um we talked to a lot of writers that hesitant, they wanted to see the platform, what it was um as we talked to them and they showed us their workflow, most of them are using grammarly um and the A I in their workflow already then as we get them on the platform and they see how it actually works and how we're looking to engage with them. Um We aren't looking to do the work for them and we're looking to be there to be helpful in that process if they need help brainstorming ideas, if they're struggling and, you know, they hit writer's block and they're looking for, where can I go with this? Where can I take it? Um So we're looking to help them through that process, but they also have the opportunity to use it as much or as little as they want. And that's one of the things that was really important for us. Um, both on the writers and the illustrator side is making sure that they have the opportunity to opt out if they want to, um, or engage when they do want to. Fantastic Colin, thank you so much for joining Fortune Brainstorm A I Yeah, absolutely. Thanks for having me. Ok.

Brainstorm AI San Francisco 2024

Brainstorm AI 2024: How Much Should We Regulate AI

Latest video

Martin Casado, General Partner, Andreessen Horowitz, Dawn Song, Professor, Computer Science; Co-director, The Berkeley Center for Responsible, Decentralized Intelligence, University of California, Berkeley

Videos (23 videos)

Brainstorm AI 2024: How Much Should We Regulate AI

Brainstorm AI 2024: Google’s Strategies For Search

Brainstorm AI 2024: The Challenge of Defining AI Transparency

Brainstorm AI 2024: Beyond Hype - Seeking AI’s Enduring Value

Brainstorm AI 2024: Chipping Away at the Competition

Brainstorm AI 2024: A New Tool to Master The Mix

Brainstorm AI 2024: The Business of Better Storytelling

Brainstorm AI 2024: AI Music Finds Its Groove

Brainstorm AI 2024: The Art of Scaling AI

Brainstorm AI 2024: Now You’re Gliding - Independence Through AI

Brainstorm AI 2024: The Role of AI In Human-Powered Health

Brainstorm AI 2024: Improving Autism Care With Intelligence

Brainstorm AI 2024: A New Lens on Stability

Brainstorm AI 2024: The Greatest Financial Instrument of All

Brainstorm AI 2024: The Everything Agent

Brainstorm AI 2024: Building A Better Model For Creativity

Brainstorm AI 2024: OpenAI Demo

Brainstorm AI 2024: The Shape of AI to Come

Brainstorm AI 2024: The Essentials of Building AI

Brainstorm AI 2024: What it Takes to Build A Startup in the AI Era

Brainstorm AI 2024: A Galaxy Of Uses For AI In Space

Brainstorm AI 2024: Smart Supply, Smarter Store

Brainstorm AI 2024: Colin Kaepernick’s Next Revolution