LeCun's Dead End Slam: Meta LLMs Can't Reason

0:00Yan Lecun, the former head of AI over at Meta has left Meta and he is shooting darts in reverse as he's leaving lighting the place on fire. He's going off to his own new startup, but he has not slowed down his criticism of Meta and the team over there on his way out, which doesn't bode well for basically what's coming out, and Zuckerberg does not appear to be very happy about this. So today on the podcast, I'm going to be breaking down what he has said, what he's off to next, and some of the criticism and some of the struggles that Meta AI is going to have going into 2026 with this whole situation.

0:36Before we get into that, if you want to try any of the AI models that I talk about on the show, including all of the Meta Llama models, including everything from OpenAI, Anthropic, Google, Grok, you know, over 40 different models, go check out AIbox.ai. You can build no-code tools. We have a playground where you can just use all of the different AI models so you don't have to have subscriptions to all of the different platforms. Go check it out, AIbox.ai. I'll leave a link in the description. All right, let's talk about what's going on with Meta AI right now. There was a big interview basically done by the Financial Times.

1:07They were talking to Yan Lekhan, who is Meta's former chief AI scientist. He had a very, you could say, blistering account of his last few months over there at the company. he said one of the most shocking things is that meta manipulated the benchmarks for their top flagship model llama 4 and he criticized their new leadership he said that large language models as a technology are a dead end there is a lot to unpack here so who is he i guess just for like a little bit of context he's definitely one of the most influential researchers of ai he's kind of known as the godfather of deep learning. He left Meta late in 2025.

1:47He was there for over a decade and he's now launching a new startup which is called Advanced Machine Intelligence Labs and he's super excited apparently to talk about why he left Meta. So one of the big things I think at the center of all of the follow is Llama 4 which is Meta's flagship language model. It was released in April last year so I mean theoretically they should have something new coming soon but Lacan acknowledged that the model's benchmarks were actually like they basically fudged the numbers on it. He said, quote, results were fudged a little bit. He then said that Meta used different model variants across benchmarks to inflate the performance, basically using whatever was the best number and calling that the final number.

2:28When this was discovered, apparently Mark Zuckerberg was not very happy. According to Yan, he said that Mark was really upset and basically lost confidence in everyone who was involved in this. According to him, Mark Zuckerberg then went and decided to sideline Meta's entire generative AI organization. He said that the consequences are still playing out. Lacan said that a lot of people have left. A lot of people who haven't left will leave. He also said that he had been largely uninvolved in Meta's LLM efforts since the original Llama model shipped, even though the company was pushing ahead with new generative models, which they codenamed Mango and Avocado, which are expected to come out this year.

3:09He said one of the funniest lines from this interview, though, he said, you don't tell a researcher what to do, especially not a researcher like me. So he obviously did not appreciate being told what to do. And part of the tension with all of that is that and part of the reason why he really was struggling with meta is because in the wake of meta acquiring scale AI, they brought in their CEO, which is Alexander Wang. He is a billionaire co-founder of Scale AI. They hired him after acquiring a 49 % stake in his company. They put him in charge of the new Frontier AI unit, which was known as the TBD Labs.

3:47And this essentially made him Yan Lekhan's boss. And Lekhan was not impressed by that. He said that he was young and inexperienced. And he said that while Wang learns quickly, he lacks experience in fundamental research and in understanding what attracts or repels top scientists. and yeah he's just you know he's he's like you certainly don't tell a researcher what to do you certainly don't tell a researcher like me what to do so i think he's just rubbed the wrong way by having a boss and not being at the top of the organization meta's broader strategy of basically giving these really huge compensation packages um you know they have reportedly had signing bonuses that were a hundred million dollars to poach researchers from open ai and others have been And, you know, a lot of people have been very skeptical of them.

4:32Lacan said the future will say whether that was a good idea or not. So, I mean, he basically confirms that this was happening. I think the conflict is a lot deeper than just the org charts and maybe Lacan's bruised ego. However, he believes that large language models are fundamentally flawed and that they're incapable of delivering true super intelligence. Inside Meta, he said that that view he had put him increasingly at odds with the leadership and with new hires who are, you know, completely LLM-pilled, according to him. He said LLMs basically are a dead end when it comes to superintelligence.

5:05Essentially, his claim is that there were efforts inside of Meta to pressure him into softening that message, but he refused. He said, I'm not going to change my mind because some dude thinks I'm wrong. I'm not wrong. My integrity as a scientist cannot allow me to do this. He also described this kind of growing disconnect between Meta's research and product team. So while the researchers are, you know, proposing riskier, more ambitious ideas, product leaders were always favoring the safer incremental bets. I mean, I think this is kind of true in a lot of organizations. He said when you do this, you fall behind.

5:35And then he also said that Meta became politically difficult. I think this kind of is because Jan LeCun has been very famously quite liberal. and Mark Zuckerberg, while he's basically supported a lot of liberal politics in the past, is now going to meetings and talking to Trump about how he's going to invest in the United States. And so I think Lacan didn't really like that shifting dynamic that he felt. Lacan's next act, the next company that he's going to be working on, is called Advanced Machine Intelligence Labs, where he's going to serve as the executive chairman. It is led day to day by their CEO, Alex LeBrun.

6:11And then Lacan says that, you know, he doesn't really want to be the CEO, but he does want to work inside the organization. It's kind of funny when he was talking about his new role at the company and why he wasn't the CEO, he said, I'm too disorganized for this and also too old. He's kind of joking, but he said his job now is to inspire. So the startup is focused on world models. This is basically an alternative approach to AI that moves. It's not just like text-only training. It's built around Lacan's VJPA architecture. So these systems learn from video and then spatial data. and physical interactions.

6:47And essentially, they're trying to understand how the world works rather than just predict a word in a model, which, honestly, this is definitely the future of AI. It's a much more complex approach. He argues that his approach avoids a lot of the structural problems of LLMs, which is, you know, hallucinations or non-deterministic reasoning, poor handling of continuous multimodal data. Early prototype systems could arrive within a year, so this isn't going to take too long. Of course, more capable versions are going to come next year. But something that's interesting he said is, he said maybe there is an obstacle we're not seeing yet, but at least there is hope.

7:21So the company has a really big footprint. It has a lot of strong ties to France. French President Emmanuel Macron apparently texted Lacan after he heard that he was leaving Meta. For now, I think one thing's pretty clear, and that is that Lacan is no longer getting kind of constrained by corporate politics. He has no intention of going quietly, So he's going to say everything that he disliked about meta. And Mark Zuckerberg definitely has a bit of a PR nightmare, I would say, on his hands. And beyond that, I think just organizational problem. There is a, you know, meta has done some interesting things with meta AI, but it definitely has a lot of fighting ahead as it's definitely had a lot of struggles.

8:00If you could do me a huge favor, if you haven't already, I would really appreciate if you could leave a review for the podcast. it helps the show out a ton and it would mean the world to me thanks so much for tuning in and i hope you have an incredible rest of your day as always make sure to check out aibox.ai it is linked in the description it's my own startup if you want to use ai to build ai tools without knowing how to code all right catch you next episode