Who is Yann LeCun? | ITPro

Geoffery Hinton — considered one of the godfathers of deep learning — quit a job at Google to speak out about the dangers of AI, but his former student and co-winner of the 2018 Turing Award holds a different opinion about AI: Yann LeCun thinks it’s dumber than a cat.

And that irreverent speech and more optimistic view on AI has become LeCun’s trademark, alongside his expertise in computer vision and convolutional neural networks (CNNs), of course.

Currently the chief AI scientist at Meta, LeCun was born in 1960 in Soisy-sous-Montmorency, a small town north of Paris in France, and became intrigued by AI after seeing 2001: A Space Odyssey. After a PhD in computer science from what is now known as the Sorbonne University with work on back-propagation algorithms, he completed a post-doc at the University of Toronto under Hinton, sparking a life-long friendship.

From there, LeCun joined the famed AT&T Bell Laboratories, developing key technologies including CNNs and image compression, and in 2003 became a professor at New York University, later founding the NYU Center for Data Science. In 2013, he joined Meta as chief AI scientist for Facebook AI Research (FAIR).

No worries about AI

LeCun and Hinton won the 2018 Turing Award together for their work on deep neural networks, alongside Yoshua Bengio. The three academics teamed up to win funding from the Canadian government to keep research moving on neural networks at a time when the field struggled to win attention — as difficult as that may seem to believe now — outlasting skepticism surrounding the technique to build systems that power the multi-billion dollar AI market today.

Though the trio share that award, LeCun, Bengio and Hinton have different opinions on the current state of AI — and the dangers raised by the technology. When stepping away from Google, Hinton said he had some regrets about his own work, later adding some of the looming dangers were “quite scary”, while Bengio was among other tech leaders calling for a pause on AI development and has argued AI safety must be taken seriously.

On the other hand, LeCun has repeatedly expressed skepticism about such existential fears, calling it “fear mongering” designed to “perform a regulatory capture of the AI industry”. He told the Wall Street Journal last year that the potential and risks of AI being a meaningful threat to us humans is “complete BS”, describing AI as dumber than a cat for the lack of ability to hold a mental model of the world or reason and plan. Criticising a UK government AI summit in 2023 that was set to focus on existential risks, LeCun posted on social media: “The UK Prime Minister has caught the Existential Fatalistic Risk from AI Delusion disease (EFRAID).”

Snark aside, LeCun says he just doesn’t believe scaling up existing techniques will be enough to gain true intelligence. “We’re missing something big to get machines to learn efficiently, like humans and animals do. We don’t know what it is yet,” he said in an interview. “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.”

Eye on open source

While LeCun didn’t sign that six-month pause letter like his Turing co-winners, he did take part in an open letter to then-US president Joe Biden, warning that AI shouldn’t “be under the control of a select few corporate entities.”

That may sound odd coming from someone employed by Meta, but under LeCun’s leadership, the Facebook-owner has taken a more open approach than closed-off rival AI developers, though critics have argued it’s not fully open source.

Indeed, LeCun has frequently spoken in favour of open source, arguing that it’s the one way to keep AI safe. Regarding DeepSeek — the Chinese developer that shocked American AI giants — he said that milestone shouldn’t be seen as China surpassing the US, but as a win for open source. “The correct reading is: ‘open-source models are surpassing proprietary ones’,” he wrote on LinkedIn.

That leaves LeCun as the pro-open-source, AI optimist attempting to cut through the hype with witticisms — regardless of whether you agree with his take on AI, at least he’s keeping the debate interesting.