Elon Musk, CEO of X, Tesla, and many other companies, and Meta AI chief Yann LeCun are currently engaged in a heated debate on X about the current capabilities and short-term potential of AI. Musk predicts that AI will be smarter than any individual human by 2024, and smarter than humanity as a whole by 2029; LeCun, for his part, pushes back by pointing out that, despite millions of training data points, we don’t even have reliable autonomous cars yet.
Why it matters. This is more than just a spat between competing egos. The dispute reflects two opposing views regarding the pace of AI, its risks, and the responsibility of informing the public about its progress. Musk, an AI evangelist, tends to exaggerate its capabilities and stoke fears of a vengeful AI. Not too long ago, he predicted that AI is, indeed, going to take all our jobs. LeCun, as a leading scientist, calls for caution and believes that we are still far from a human-level AI, criticizing alarmist claims. In the past, he had already labeled the threat of AI to humanity as “completely ridiculous.”
Contrasting positions.
- Musk has warned of a small probability that AI could “annihilate humanity.” At one point he called for a six-month halt on all AI development (coincidentally just before launching his own AI project).
- LeCun argues that AI has the computing power of a cat, but nowhere near its intelligence. He sticks to the car analogy: If Musk were right, we’d already have AI systems capable of learning to drive in 20 hours, just like a teenager. This, however, is not the case.
Background. This isn’t the first time both men have clashed. Musk has a well-known history of belligerence and sarcasm. And although LeCun has a more academic and scientific profile, he never shies away from confrontation, and has mocked OpenAI in the past (Musk was one of OpenAI’s founders).
Musk has been proclaiming the imminent arrival of Tesla’s robotaxis for years, but the deadlines keep being pushed back. Recently, he raised $6 billion for xAI, his new AI startup. To recruit talent, he called upon scientists who were committed to “understanding the universe” and “the pursuit of truth, without regard to popularity or political correctness.”
The reply. LeCun responded harshly: “Join xAI if you can stand a boss who: Claims that what you are working on will be solved next year (no pressure). Claims that what you are working on will kill everyone... Claims to want a ‘maximally rigorous pursuit of the truth’ but spews crazy-ass conspiracy theories on his own social platform.”
Meanwhile, Musk accused LeCun of “going soft.” In response to LeCun stating that he's published over 80 technical papers since January 2022, Musk minimized the achievement and told him: “Try harder!”
So, who’s right? Musk has a history of broken promises regarding AI, but also undeniable achievements with companies like Tesla, SpaceX, and now Neuralink. However, he's also known for his outbursts and conspiracy theories. LeCun, on the other hand, is a leading AI scientist, which gives him authority to speak on the complex subject of AI and a realistic view of AI’s current limits that Musk lacks. Still, the last five years have seen such significant advancements in AI that almost any medium-term possibility seems plausible.
Image | Xataka, Wikimedia Commons
Related | GPT-4o: What Is It and How Is It Different From GPT-4?