I like this post. It reminds me of contemporary programming language research. There are precious few people actually doing interesting stuff in PL research these days that are actually trying to uncover new paradigms that aren't just the same old "we discovered how to do X in Y type system" or "novel technique to generate objects with blah blah constraints".

People doing actually interesting stuff can't get funding, so they have to lone-wolf their entire research or just give up and work on stuff that gets paid, people like

- Jonathan Edwards

- Allen Webster

- Brett Victor

All with seriously intriguing ideas that probably have potential, but nobody seems to want to actually dig in to the stuff. Fortunately, there are guys like Stephen Kell who are kind of doing it even in academia, but I think he's limited too towards working on the boring problems that get funding as well.