The speaker believes it is an overcorrection to regard all work outside machine-learning research as worthless.

“That's prestigious, and that continues to be prestigious. But I think the overcorrection that people have is that looking at everything else that is not that ML research, that is worthless.”