Name: Insights: The Promise AI Holds for Business
Uploaded: 2020-02-25
Duration: 9 min 37 s
Description: Fortune's Brainstorm Films explores the wonders and dangers of Artificial Intelligence. (October 2018)

[MUSIC PLAYING] ADAM LASHINSKY: Now, AI is being incorporated into all sorts of things. The Alexa Echo, or Google Home, or Siri, that's very garden variety AI. This is really first inning AI. MARY BETH LAUGHTON: We always try to not just do technology for technology's sake, but actually solving a consumer problem pain point. We feel like we need to have the right data and the right talent to actually really take advantage of AI, so we're kind of focused on that first. But we're also experimenting in a couple different ways of getting some AI out there just to test. ANDREW NUSCA: Most CEOs are asking themselves what can I do with AI? How can I incorporate this? This is clearly the future, but they don't really know. And that's part of the problem. It's very abstract. Most of the applications you see today involve machine learning. Google image search, that's machine learning comparing things. The certain kind of ad on Facebook, that's machine learning trying to tailor those ads to the preferences that you have expressed on Facebook. The more advanced stuff, that's not really in the wild yet but for the largest tech companies. NAVEEN RAO: So artificial intelligence actually has its roots back in computing-- Alan Turing even. He was really trying to recapitulate pieces of intelligence. What made a machine intelligent? And he recognized very early on that, that lived in the information space. Processing information to make it useful somehow. The term artificial intelligence was coined in the '60s. We started to understand the brain was definitely electrical. In the '40s, we actually defined how it works electrically, how information is represented, what we call spiking. That actually was concurrent with a lot of innovations on the computing side. Because we said, well, brains are information tools, computers are information tools, and there's always this parallel. The first computer science department and the first neuroscience department, started a year apart. And we figured out that neurons represent information a certain way. We tried to simulate those things in synthetic systems. We thought it would be really easy because we figured out all these principles from the brain. It turns out it's a little more difficult than that. So the field kind of had its ups and downs until computing got to a point where it was dense enough. And the data set sizes got to the point where they're big enough. 2010, 2011 time frame is when neural networks met Moore's law met big data. This whole field exploded. It's going to change with time, but overall the goal of a learning system is to find useful structure and data. ADAM LASHINSKY: Prominent AI researchers have said that AI will be like electricity. Electricity was a big deal. It was this new invention that people talked about. Nobody talks about electricity anymore and that's where we are with AI. X years down the road, sooner rather than later, AI will be infused in all these things in ways that we couldn't possibly fathom today. ANN CAIRNS: The way that you use a telephone is different from the way that I use a telephone, the way that we actually key. We have artificial intelligence that can detect your behavior. So if I left my phone there and you picked it up and tried to make a payment, it would know it's not me. JEREMY KING: When we're unloading a truck full of strawberries, one of the drivers can open up the case and take a photo. We can determine whether it was sent sustainably and how long that fruit is going to last. MARY BETH LAUGHTON: I think AI can be the future of personalization. And I think personalization is where retail and beauty are going. We have what's called Sephora Virtual Artist, which is our augmented reality beauty experience. We're using computer vision, so it can map your facial features and know where to place the makeup when you look at it virtually. SPEAKER 1: We've actually been using AI to really reshape the entire end-to-end shopping journey for the customer. Increasingly, we use AI to show up through targeted display ads in places where the consumer is shopping. So if you let us know that you're interested in certain lifestyles and certain items, we'll make sure that we personalize the recommendations that we make for you. CHRISTOPHER HEISER: Our customers use a lot of AI. The self-driving systems use it for object detection. We work with people that are perceiving the occupants. Are they happy? Are they scared? Are losing trust with the vehicle? There are already dozens of applications of AI. I only expect that to increase. ANDREW NUSCA: There is no technological development with more potential than artificial intelligence. Because it is a suite of capabilities, it's going to solve simple things in our lives and the extraordinary things. What is really exciting about that for the business community, there's money to be made. MICHAL LEV-RAM: There's so much excitement around AI, rightfully so. But if we're not thoughtful about how we go about building these systems, the outcome could be disastrous. ANN CAIRNS: If there aren't enough of us engaged, and I don't actually even just mean women, I mean a diverse group of people, then we're going to get computers who think like some bro culture on the West Coast. MICHAL LEV-RAM: Bias can be inherent to AI and to algorithms. If we allow it to be, comes down to people. If those teams aren't reflective of humanity, then the algorithms aren't going to be reflective of humanity and that has hugely negative ramifications. ANDREW NUSCA: One of the biggest concerns is that workplaces will lose their humanity if we keep incorporating more and more AI into the office. CHRISTINA CULCLOUGH: There's a lot of algorithms that are being put into place that help the companies in their recruitment processes, who are they going to hire, who they're going to fire, and so on. But the big question is what data are they using? Are they bringing in your social media data, your private data? Do they get access to your whole file, your financial data? CHRISTOPHER HEISER: The amount of information that we freely give or unknowingly give to corporations and governments is massive. AI will allow those groups to manage in a very efficient way. Whether that's something that's good for me or bad for me, I think, depends on who's doing it and how. CHRISTINA CULCLOUGH: We have all sleepwalked into this current situation. We haven't really understood what it was we were giving away when we were using all of these services, when we're using social media. What is it we are paying with to get all of these services? CHRISTOPHER HEISER: We've got people arguing that AI is great and other people arguing that it's the end of humanity. The answer is always somewhere in between. It is something we have to think pretty deeply about. CHRISTINA CULCLOUGH: Every single technological development has changed jobs. I mean, that's nothing new. Where I think this whole transformation is new is its magnitude. It's going to affect all of us. ANDREW NG: I am concerned that these hype discussion about evil AI, killer robots, it whitewashes the real challenges. And I think job displacement is a big one. CHRISTINA CULCLOUGH: A lot of jobs will change, meaning that we will have to learn new things. And this brings us to the whole question of re and upskilling. This is going to be one of the major challenges of our time, making sure that everybody can follow the developments and have an opportunity to thrive in the labor market. JEREMY KING: What happens is jobs shift? We have about 18,000 new jobs just for grocery picking and delivery services. JUSTIN TRUDEAU: Robotics, AI, technological advances are rapidly disrupting and transforming the nature of work and the workplace. We know what happens with industrial revolutions. There is a transition time in which a whole bunch of people get disrupted right out of their jobs. Once it's through that transition time, there are better quality jobs and more jobs for everyone. CHRISTINA CULCLOUGH: Predictions are that AI will create many more jobs than it will destroy. If those jobs that will be created are bad jobs, are paid under their living wage, are precarious, then that is nothing we should celebrate. JEREMY KING: I don't know if you've seen this picture of sheep problem. If you showed a computer scientist, three or four years ago, a picture of a flock of sheep and ask them to build an algorithm to count how many sheep were in the picture, they would have told you it's impossible. Now, that's trivial. Overnight that changed. When I think about how fast AI, machine learning is going to move and that's the kind of speed we're looking at, that's exciting. NAVEEN RAO: Right now the depth of expertise required is extremely high. CHRISTOPHER HEISER: It's not integrated at all in the way the people do their business today. MARY BETH LOUGHTON: It's a revolutionary technology. CHRISTOPHER HEISER: Certainly, the machines are getting much, much better at a lot of things than we are. MICHAL LEV-RAM: It's going to be much more integrated into our lives. CHRISTOPHER HEISER: I think we're just at the beginning of trying to understand how it fits in. [MUSIC PLAYING] 

Insights: The Promise AI Holds for Business

Uncategorized

The wonders and dangers of Artificial Intelligence.

The Promise AI Holds for Business

Fortune's Brainstorm Films explores the wonders and dangers of Artificial Intelligence. (October 2018)

now May I being incorporated into all sorts of things the Alexa Echo or Google Home or Serie? That's very garden variety. This is really first inning way. Always try to not just technology for technology's sake, but actually solving a consumer problem or main point. We feel like we need to have the right data and the right talent actually really take advantage of a I. So we're kind of focused on that, first, experimenting in a couple of different ways of getting some aye aye out there. Most CEOs were asking themselves, What can I do with a I? How can I incorporate this? This is clearly the future, but they don't really know. And that's part of the problem. It's very abstract. Most of the applications you see today involved machine learning, Google Image search, right, That's machine learning. Comparing a certain kind of add on Facebook that machine learning trying to tailor those adds to the preference is that you have expressed on Facebook the more advanced stuff that's not really in the wild yet, but for the largest tech, Artificial intelligence actually has its roots back in computing. Alan Turing, even he was really trying to recapitulate pieces of intelligence, what made a machine intelligence. And he recognized very early on that that lived in the information processing information to make it useful. Somehow term artificial intelligence was coined. In the sixties. We started to understand the brain was definitely electrical in the forties. We actually defined how it works electorally, how information is represented, what we call spiking that actually was concurrent with a lot of innovations on computing side because, he said, Well, Marines are information tools, computers, air information always parallel. First computer science department At First Neuroscience Department started a year apart. We figured out that neurons represent information certain way we tried to simulate those things and synthetic systems. We thought it would be really easy because he figured out all these principles. Brain turns out a little more difficult. So the field kind of had its ups and downs until computing got to a point where it was dense enough and the data set size has got to the point where they're big enough. 10 4011 timeframe. When neural networks met, Moore's law met big data. This whole field exploded. It's gonna change with time But overall, the goal of ah warning system is to find useful structure in data prominent Ai ai. Researchers have said that a I will be like electricity. Electricity was a big deal. It was this new invention that people talked about. Nobody talks about electricity anymore. And that's where we are with a I X years down the road sooner rather than later, eh? I will be in used in all these things in ways that we couldn't possibly battle today. The way that you used a telephone is different from the way that I used the telephone. The way that we actually came, we have artificial intelligence that can detect your behavior. So if I left my phone there and you picked it up and try to make a payment, it would know not me. When we're unloading a truck full of strawberries, one of the drivers and open up the case and take a photo waken determine whether it was sent sustainably. How long that food is gonna last. I think a I can be the future of personalization, and I think personalization is where retail and you're going. We have what's called for virtual artists which is our augmented reality beauty experience. We're using computer vision so it can kind of map your facial features and know where to place the makeup When you look at it. Virtually, we've actually been using a I really re shape the entire into end shopping journey for the customer increasingly used a I to show up through targeted display at places where the consumers shopping. So if you let us know that you're interested in certain lifestyles and certain items, we'll make sure that we personalize the recommendations that we make for you. Our customers use a lot of self driving systems. Music for object detection way work with people that are receiving the occupants. Are they happy or they're scared of losing trust. There are already dozens of application today. There is no technological development with more potential than artificially because it is a suite of capabilities. It's gonna solve simple things in our lives and the extraordinary What is really exciting about that business community? There's money to be made. There's so much excitement around, eh? I rightfully so. But if we're not thoughtful about how we go about building these systems, the outcome could be disastrous. if there aren't enough of us engaged and I don't actually even just mean women. I mean, a diverse group of people. Then we're gonna get computers who think like some bro culture on the West Coast bias ca NBI inherent to a i into algorithms if we allow it to be come down to people with those teams aren't reflective of humanity than the algorithms aren't going to be reflective of humanity. And that has hugely negative ramifications in one of the biggest concerns is that workplaces will lose their humanity if we keep incorporating more and more A i into the office. There's a lot of algorithms that are being put into place that helped the companies in the recruitment process is who they're gonna hire, who they're going to fire on so want. But the big question is what data they're using. Are they bringing in your social media data? Your private data? Do they get access to your health file your financial data, the amount of information that we freely give or unknowingly give to corporations and governments Massive, eh? I will allow those groups to manage in a very efficient way, Whether that's something that's good for me or bad for me and think depends on who's doing it and how we have all sleep walked into this current situation. We haven't really understood what itwas we were giving away. When we're using all of these service is when we're using social media. What is it we are paying with? Get a load of these. We've got people arguing that they eyes great and other people are doing the end of humanity. The answer is always somewhere in between it It's something we have to think deeply about. Every single technological development has changed jobs, right? I mean, that's nothing. You where I think this whole transformation is new, is its magnitude. It's gonna affect a lot of us. E. I'm concerned that these high discussion about evil robots it whitewashes the real challenges. And I think this is a big one. A lot off jobs would change, meaning that we will have to learn new things. And this brings us to the whole question off re and ups. Killing this is gonna be one of the major challenges of our time, making sure that everybody can follow the developments and have an opportunity to live in the labor market. What happens is jobs shift right. We have about 18,000 new jobs just for grocery picking and delivery service. Robotic. Aye, aye. Technological advances are rapidly disrupting and transforming the nature of work and the workplace. We know what happens with industrial Revolution. There is a transition time in which the whole bunch of people get disrupted right out of their jobs. Once it's through that transition time, there are better quality jobs and more jobs for everyone. Predictions are a I will create many more jobs than it will destroy. If those jobs that will be created are bad, jobs are paid under the living wage, a precarious, then that is nothing we should celebrate. I don't know if you've seen this picture of sheep problem. You showed a computer scientist three or four years ago a picture of a flock of sheep and asked them to build an algorithm to count how many people were in the picture, they would have told you is impossible. Now that's trivial. Overnight that change when I think about how fast a I machine learning is gonna move. And that's the kind of speed we're looking at. That's exciting. Right now. The depth of expertise required is extremely high. It's not integrated at all with people today. Revolutionary technology. Certainly the machines are getting much, much better. A lot of things. Yeah, I'm going to be much more integrated into our lives. I think we're just at the beginning of trying to understand.