August 2024: "generative AI is a dead-end technology that has peaked”
July 2024: "Generative AI models aren’t getting more energy-efficient, nor are they getting more “powerful” in a way that would increase their functionality"
Nov 2025: "the fact we're running out of high quality training data and we're hitting the walls of scaling laws, in the training paradigm, these models aren't getting better. What we're seeing today is pretty much what they're always gonna be like"
There are other examples too from his prose of talking about how they are barely useful but I don't want to dig it up
He does say that pretty often in interviews. That doesn’t change his thesis but that‘s one reasonable reason people dismiss him, he has often said that AI is useless when considering the externalities. And he will sometimes take a shortcut and just say „it’s useless, doesn’t do anything well“, which is of course way too simplified.