How Not to Be Stupid About AI, With Yann LeCun
It’ll take over the world. It won’t subjugate humans. For Meta’s chief AI scientist, both things are true.
BY STEVEN LEVY
BACKCHANNEL
DEC 22, 2023 6:00 AM
How Not to Be Stupid About AI, With Yann LeCun
It’ll take over the world. It won’t subjugate humans. For Meta’s chief AI scientist, both things are true.
PHOTOGRAPH: ERIK TANNERBACKCHANNEL
DEC 22, 2023 6:00 AM
How Not to Be Stupid About AI, With Yann LeCun
It’ll take over the world. It won’t subjugate humans. For Meta’s chief AI scientist, both things are true.
DO NOT PREACH doom to Yann LeCun. A pioneer of modern AI and Meta’s chief AI scientist, LeCun is one of the technology’s most vocal defenders. He scoffs at his peers’ dystopian scenarios of supercharged misinformation and even, eventually, human extinction. He’s known to fire off a vicious tweet (or whatever they’re called in the land of X) to call out the fearmongers. When his former collaborators Geoffrey Hinton and Yoshua Bengio put their names at the top of a statement calling AI a “societal-scale risk,” LeCun stayed away. Instead, he signed an open letter to US president Joe Biden urging an embrace of open source AI and declaring that it “should not be under the control of a select few corporate entities.”
LeCun’s views matter. Along with Hinton and Bengio, he helped create the deep learning approach that’s been critical to leveling up AI—work for which the trio later earned the Turing Award, computing’s highest honor. Meta scored a major coup when the company (then Facebook) recruited him to be founding director of the Facebook AI Research lab (FAIR) in 2013. He’s also a professor at NYU. More recently, he helped persuade CEO Mark Zuckerberg to share some of Meta’s AI technology with the world: This summer, the company launched an open source large language model called Llama 2, which competes with LLMs from OpenAI, Microsoft, and Google—the “select few corporate entities” implied in the letter to Biden. Critics warn that this open source strategy might allow bad actors to make changes to the code and remove guardrails that minimize racist garbage and other toxic output from LLMs; LeCun, AI’s most prominent Pangloss, thinks humanity can deal with it.
I sat down with LeCun in a conference room at Meta’s Midtown office in New York City this fall. We talked about open source, why he thinks AI danger is overhyped, and whether a computer could move the human heart the way a Charlie Parker sax solo can. (LeCun, who grew up just outside Paris, frequently haunts the jazz clubs of NYC.) We followed up with another conversation in December, while LeCun attended the influential annual NeurIPS conference in New Orleans—a conference where he is regarded as a god. The interview has been edited for length and clarity.
Steven Levy: In a recent talk, you said, “Machine learning sucks.” Why would an AI pioneer like you say that?
Yann LeCun: Machine learning is great. But the idea that somehow we're going to just scale up the techniques that we have and get to human-level AI? No. We're missing something big to get machines to learn efficiently, like humans and animals do. We don't know what it is yet.
I don't want to bash those systems or say they’re useless—I spent my career working on them. But we have to dampen the excitement some people have that we're just going to scale this up and pretty soon we’re gonna get human intelligence. Absolutely not.
You act as though it’s your duty to call this stuff out.
Yeah. AI will bring a lot of benefits to the world. But people are exploiting the fear about the technology, and we’re running the risk of scaring people away from it. That's a mistake we made with other technologies that revolutionized the world. Take the invention of the printing press in the 15th century. The Catholic Church hated it, right? People were going to be able to read the Bible themselves and not talk to the priest. Pretty much all the establishment was against the wide use of the printing press because it would change the power structure. They were right—it created 200 years of religious conflict. But it also brought about the Enlightenment. [Note: Historians might point out that the Church actually made use of the printing press for its own purposes, but whatever.]
Why are so many prominent people in tech sounding the alarm on AI?
Some people are seeking attention, other people are naive about what's really going on today. They don't realize that AI actually mitigates dangers like hate speech, misinformation, propagandist attempts to corrupt the electoral system. At Meta we’ve had enormous progress using AI for things like that. Five years ago, of all the hate speech that Facebook removed from the platform, about 20 to 25 percent was taken down preemptively by AI systems before anybody saw it. Last year, it was 95 percent.
How do you view chatbots? Are they powerful enough to displace human jobs?
They’re amazing. Big progress. They’re going to democratize creativity to some extent. They can produce very fluent text with very good style. But they’re boring, and what they come up with can be completely false.
“AI will bring a lot of benefits to the world. But we’re running the risk of scaring people away from it.”
The company you work for seems pretty hell bent on developing them and putting them into products.
There's a long-term future in which absolutely all of our interactions with the digital world—and, to some extent, with each other—will be mediated by AI systems. We have to experiment with things that are not powerful enough to do this right now, but are on the way to that. Like chatbots that you can talk to on WhatsApp. Or that help you in your daily life and help you create stuff, whether it's text or translation in real time, things like that. Or in the metaverse possibly.
How involved is Mark Zuckerberg in Meta’s AI push?
Mark is very much involved. I had a discussion with Mark early in the year and told him what I just told you, that there is a future in which all our interactions will be mediated by AI. ChatGPT showed us that AI could be useful for new products sooner than we anticipated. We saw that the public was much more captivated by the capabilities than we thought they would be. So Mark made the decision to create a product division focused on generative AI.
Why did Meta decide that Llama code would be shared with others, open source style?
When you have an open platform that a lot of people can contribute to, progress becomes faster. The systems you end up with are more secure and perform better. Imagine a future in which all of our interactions with the digital world are mediated by an AI system. You do not want that AI system to be controlled by a small number of companies on the West Coast of the US. Maybe the Americans won't care, maybe the American government won't care. But I tell you right now, in Europe, they won't like it. They say, “OK, well, this speaks English correctly. But what about French? What about German? What about Hungarian? Or Dutch or whatever? What did you train it on? How does that reflect our culture?”
Seems like a good way to get startups to use your product and kneecap your competitors.
We don’t need to kneecap anyone. This is the way the world is going to go. AI has to be open source because we need a common infrastructure when a platform is becoming an essential part of the fabric of communication.
