You can't. It's AI booster propaganda. Yann LeCun obviously understands transformers better than random anons on this venture capital fan forum. This means he also understands their limitations, which is too complex for people who just want their preferred billionaire to win.

> "I take an object, I put it on the table, and I push the table... there's no text in the world, I believe, that explains this... your GPT-5000, or whatever it is, it's never gonna learn about this."

Lex fridman podcast 2022

ChatGPT answers this just fine.

He also said they will be “ Writing aids, not much more" and they “Can't plan.”

No one said he knows nothing about building AI. He was just clearly wrong about LLM capabilities.

>> "I take an object, I put it on the table, and I push the table... there's no text in the world, I believe, that explains this... your GPT-5000, or whatever it is, it's never gonna learn about this."

As far as I know, those things are still true. Unless there is text in the training material that the text in the prompt can map to, or there is text in training material that describe how to map text in the prompt to text in the training material, LLMs cannot process the prompt.

So may be he is wrong about what could be in the training material. But I don't think he is wrong in the fundamental..

This is sort of what I mention in my original comment. These claims, when pushed, turns out not to hold water...

Yann is smart but obviously not omniscient, I don't think it does the ML community a service to pretend that elders are never incorrect (or even badly incorrect). This type of things is how we got the cult around Schmidhubering.