In the debate over artificial intelligence, whose side is Elon Musk on?
Musk, who’s in charge of SpaceX, Tesla and the Neuralink brain interface venture, sized up the odds with AliBaba founder Jack Ma today during a widely watched one-on-one session at the World Artificial Intelligence Conference in Shanghai.
The way Musk sees it, the prospects aren’t great for humans if future AI agents decide to go rogue. That’s despite pronouncements from AI researchers who say machines won’t match humans anytime soon when it comes to general intelligence, as opposed to specialized AI applications such as playing chess or Go.
“The biggest mistake I see artificial intelligence researchers making is assuming that they’re intelligent,” Musk said. “Yeah, they’re not, compared to AI. And so a lot of them cannot imagine something smarter than themselves.”
Musk said future AI agents will be “vastly smarter” than humans. “So what do you do with a situation like that?” he asked. “I’m not sure. I hope they’re nice.”
For his part, Ma saw more promise than peril in AI.
“I’m not a tech guy. I think more about life,” Ma said. “I think AI is going to open a new chapter of the society of the world [in which] people try to understand ourselves better than the outside world.”
Ma said he wasn’t concerned about machines replacing humans in the workplace. “I think that because of artificial intelligence, people will have more time to enjoy human beings. I don’t think we’ll need a lot of jobs,” he said.
He went so far as to say that in the next 20 or 30 years, people could get by working three days a week, four hours a day. (After the debate, The Washington Post noted that Ma’s prediction struck some Chinese observers as grimly ironic, in light of his past praise of a 72-hour weekly work scheme that’s known as “996” because it involves working from 9 a.m. to 9 p.m. six days a week.)
Ma implied that humans would be able to keep the machines under control. “We invented a computer,” he said. “I never saw a computer invent a human being.”
Musk was far less sanguine. When Ma said “street-smart” people like himself will never be scared of AI and will always think it’s great fun, Musk replied: “I don’t know, man, that’s like ‘famous last words.’ ”
That’s why Musk is putting so much money into Neuralink, which recently reported progress in its effort to create an integrated brain-machine interface.
“If you can’t beat ’em, join ’em,” Musk said. “That’s what Neuralink is about. Can we go along for the ride with AI?”
Other memorable moments from the Musk-Ma meetup:
- Ma joked that he wasn’t crazy about the term “artificial intelligence,” even though his company has a network of AI labs around the world, including in Seattle. “I call it ‘AliBaba Intelligence,” he said.
- Ma took aim at Musk’s plan to send settlers to Mars on the Starship launch system he’s developing. “It’s great to send 1 million people to Mars, but we have to care about the 7.4 billion people on Earth,” he said. “How can we make the world more sustainable?” Musk replied that his efforts to advance clean energy with Tesla could make a big difference in keeping Earth sustainable, and that his vision for Mars was meant merely to ensure that humanity survives as a multiplanetary species, even if Earth suffers a catastrophe. “To be clear, I’m pro-Earth,” he said.
- Musk said it looks as if it’s up to humanity to keep the flame of consciousness burning, at least in our corner of the universe. “There are people out there who think we have found aliens,” he said. “Trust me, I would know. We have not. People ask me if I’ve been to Area 51. OK, please … SpaceX actually has Area 59 — eight better than Area 51.”
- Doesn’t the prowess of game-playing programs such as Google’s AlphaGo prove AI’s superiority? Ma said it would be “stupid” to say machines were smarter than humans on that basis, just as it would be wrong to dwell on an automobile’s ability to go faster than a human. He contended that competitions like track and field, or Go and chess, are meant to pit humans vs. humans, not humans vs. machines. “I never, ever play chess or Go with a computer,” Ma said.
- Musk noted that he was visiting Shanghai in part to visit Tesla’s Gigafactory 3 in China. Some media outlets speculated that Musk would show off the first Tesla electric car produced at the Chinese factory. Musk was also due to launch the Chinese division of his Boring Company tunneling venture.
- Musk and Ma agreed that it’s hard to predict the future. Ma said “99.99% of the predictions that human beings had in history about the future — all wrong.” Which led Musk to ask, “Including that one?”