AI -> Writing Lisp Books -> First Startup -> Designing Arc -> YC -> Designing Bel (for 4+ years)
That is, he was interested in Lisp before he made a lot of money, and also after he made a lot of money (twice). The money didn't change his interests.
Writing your own Lisp isn't very likely to lead to "concrete" results, but that's the point (or the lesson). Someone trying to do only "practical" things wouldn't have created a valuable startup or YC in the first place.
There are a few commenters here who missed that point, assuming that everyone must be doing things to get rich (because that's likely their own motivation).