Comment by anothername12
7 months ago
Favorite bit:
> very difficult to find much on Raghavan’s history [..] but from what I’ve gleaned, his expertise lies primarily in “failing up,”
7 months ago
Favorite bit:
> very difficult to find much on Raghavan’s history [..] but from what I’ve gleaned, his expertise lies primarily in “failing up,”
Yeah he only co-authored 2 highly influential CS textbooks when he wasn't failing upward
That doesn't mean his insights are correct, that he's a good leader, or that his efforts have been good for us users.
It's hard to find a bigger name in AI than Minsky, and he wrote one of the most influential books in Computer Science, which put a huge damper on all neutral network progress for decades... Arguably, that was a bad thing.
Neural networks put a huge damper on themselves by being so compute hungry.
They've always been there, it was just an inefficient way to do machine learning before around 2011 when GPUs started being good enough