He often says in passing that he doesn't see value in LLMs outside of generating bad software and can help sometimes with search. But it's not really part of th

He often says in passing that he doesn't see value in LLMs outside of generating bad software and can help sometimes with search. But it's not really part of the AI bubble thesis itself IMHO, just him adding his personal commentary in interviews