The kind of superintelligent AI that has prompted doomsday warnings from some of the world’s most prominent tech minds will be with us within a decade, according to SoftBank’s founder and CEO—who compared those resisting the technology to “goldfish.”
Speaking at the SoftBank World corporate conference on Wednesday, Masayoshi Son predicted artificial general intelligence (AGI)—a form of AI that will be smarter than people and able to outperform humans in a broad range of tasks—would become a reality in less than 10 years’ time.
By 2030, he insisted, AI will be 10 times as intelligent as all of humanity. This would result in societal advances like a complete transition to autonomous vehicles, as well as AI-generated advances in science and technology worthy of the Nobel Prize.
Son, who built a reputation as an industry visionary with his prescient investment in Alibaba back in 2000, urged the audience listening to him at the event to “take advantage of it or be left behind.”
“Do you want to be a goldfish?” He asked attendees, arguing that those who avoided adopting the technology would become apelike in their intellectual capabilities compared with those who embraced AI.
“Saying, ‘Don’t use AI,’ is like saying, ‘Don’t drive a car or use electricity,’” Son added. “Whether you like it or not, the AI revolution will come.”
He also labeled anyone denying AI’s potential a “hallucinator”—a term that has been used to describe mistakes made by generative artificial intelligence chatbots like OpenAI’s ChatGPT.
Son has made no secret of his, or SoftBank’s, interest in AI.
While SoftBank has wound down its investments over the past year on the back of losses driven by tech volatility, the Japanese conglomerate has laid out plans to double down on artificial intelligence.
A number of the company’s portfolio firms are involved in AI development, including British chipmaker Arm, which recently enjoyed the biggest IPO of the year.
Last week, the Financial Timesreported that OpenAI was in talks with SoftBank and ex–Apple designer Jony Ive to develop an “iPhone of artificial intelligence.”
The publication also reported that SoftBank was on the hunt for AI deals following Arm’s IPO, with Son wanting to go on a multibillion-dollar spending spree within the red-hot sector.
Will AI outsmart humans?
Since OpenAI’s large language model chatbot ChatGPT took the world by storm late last year, artificial intelligence has generated countless headlines, won billions of dollars from investors, and divided experts on how it will change the planet.
For his part, Son made it clear on Thursday that he’s a firm believer AI will be a catalyst for global transformation—and that the tech was already close to outsmarting people.
“It is wrong to say that AI cannot be smarter than humans, as it is created by humans,” he said. “AI is now self-learning, self-training, and self-inferencing, just like human beings.”
Many prominent technologists—including Tesla CEO Elon Musk, Apple cofounder Steve Wozniak, and two of the three scientists known as the “godfathers of AI”—have issued chilling warnings about the potential artificial intelligence has to destroy humanity. Other experts like Son, however, have taken a much more pro-AI stance.
Over the summer, more than 1,300 AI experts signed a letter that downplayed the notion of AI generating an army of “evil robot overlords,” arguing that the tech was “a force for good, not a threat to humanity.”
Meanwhile, LinkedIn cofounder Reid Hoffman dismissed calls for AI research to be paused for six months as “foolish” and “anti-humanist” at a conference in London last month.
Tech experts are also divided on how long it will take for AI to reach a level of intelligence that surpasses that of human beings.
Musk, who recently launched his own AI startup, has warned that artificial intelligence could be “vastly smarter” than any human as early as 2025.
However, Nick Clegg, president of global affairs at Facebook and Instagram parent firm Meta, said in July that many current iterations of the technology are still “quite stupid.”
Christina Maher, a biomedical engineer and neuroscientist at the University of Sydney, wrote in an article earlier this year that while AI would eventually reach human-level intelligence, it was unlikely this would be achieved “anytime soon.”
“What’s left is for AI models to learn inherent human traits such as critical reasoning, and understanding what emotion is and which events might prompt it,” she said. “AI hasn’t acquired these capabilities yet. But if humans can learn these traits, AI probably can, too—and maybe at an even faster rate.”
Geoffrey Hinton, one of AI’s “godfathers,” said in a May interview, however, that AI developers were “very close” to creating machines whose intelligence superseded humanity’s.
“They will be much more intelligent than us in the future,” he said. “How do we survive that?”