Transcription
Yann LeCun's critique comes down to four capabilities that LLMs structurally cannot have.
Number one, there's no model of the physical world. An LLM can describe the world in multiple languages, but it's never been to Paris or seen a waterfall or or touched a puppy. Unlike humans, there's no intuitive understanding of gravity, friction, heat, or hunger. These are just tokens which have no innate understanding. And a 4-month-old baby has a much better physics engine than GPT-5. And this means that human intelligence is much much more than just predicting the next word in a sentence.
Number two, no persistent memory. You may not realize this, but every time you interact with AI, it starts from zero. Behind the scenes, your favorite chatbot is managing context history, and it's getting injected into the conversation every time, which is why the every conversation starts with zero. In simple terms, context windows is like a goldfish with a really good notebook. But a context memory, which is what humans have, is like an elephant who never forgets being mishandled.
Number three, large language models have no real ability to reason. They're just guessing about information, but they're not actually retrieving information. They're predicting the next token. And when they appear to reason, what they're doing is pattern matching against reasoning patterns they've already been trained on through Reddit or any other source on the internet. And that's why they can solve a hard math problem from a textbook, but then fail on something trivial that a child could handle. What's most important about being human is not ever written down. And if you compare this to a human, a humans have cultural values and social norms and can make an inbuilt assessment of whether something is going to hurt them or help them. But an LLM has no skin in the game because there's no skin. It's just predicting what the right answer might look like, which is a very different thing to finding the right answer.
Number four, limited planning, and this might surprise you. If you go ahead and ask an LLM to plan a week of work for you, it will produce something that looks like a plan, but it's not actually simulating the future. It's just trying to plan your week based on existing patterns. If you show it something new, it will have you meeting a customer at one end of the city and then put the very next meeting on the other side of town. There's no inherent ability to take into account factors such as geography or time differences. It's going to give you something that looks plausible, but probably not useful. Language, LeCun argues, is a thin compressed shadow of reality. And training on text alone is like learning to swim by reading a textbook.