It's not so much they _turned_ wrong, they were obviously wrong at the time he made statements, or shortly after. He is just extremely biased against transformers, and it's easy to see why - transformers are already invented, and as a scientist he wants to invent something new.

More specifically:

* he said that autoregressive transformer is fundamentally flawed: if it makes one error, all further generated tokens make no sense; that's, obviously, bullshit: there are many ways to implement backtracking, and there are many papers demonstrating backtracking at a time he did that lecture. And, of course, it just doesn't matter - reasoning model can just write "Oh, that's wrong. Let's try a different thing". - it's just not derailed by one mistake.

* he said that GPT-1000 can't possible have understanding of physical world, when GPT-4 already existing at that time could solve a problem he described.

So it's not like he was bad at predicting things, he's intellectually dishonest.

Yeah that slide of him "proving" autoregressive is flawed was really the nail in the coffin. I stopped listening to him after that.