For what it’s worth, Scott spent the last few years working for OpenAI because he wanted to do something much more applied than quantum information theory. As an applied scientist I’m aghast: he’s one of the few people who actually gets to “touch the firmament of the Universe,” why would you ever give that up for mere applied science (even science as interesting as whatever goes on inside OpenAI) :) But as a human being I understand it. The grass is always greener somewhere else, and doing tangible things is sometimes more fun.

TL;DR: whatever you’re doing, there’s probably someone who wishes they were doing it instead.