[jerf]

You'd better hope we can figure out some higher abstractions for QM, or you can kiss the even-harder problem of biological computing goodbye.

I think there will be, though. I think 99% of the problem is just that we have no hardware yet. It's historically just been too hard to develop algorithms for things we don't have the ability to run yet. I've seen it personally in evolutionary computation [1], and the recent renaissance in AI and ML I believe was largely driven by getting to the point we could actually run these computations in some reasonable period of time. The breakthroughs probably could have happened sooner, except even if you theorized about Deep Learning, nobody would have even been able to use it.

[1]: A professor of mine told a heartbreaking (to me) story of carrying around a deck of cards between several institutions as he progressed through his early academic career, running an evolutionary computation in whatever spare time he could get over the literal years. My crappy, grad-student-grade personal laptop, a cheap piece of crap even for the time (I had to permanently clock the nominally 1GHz CPU down to 500MHz just to keep the thing from burning itself up), could have done the whole thing in minutes, if not seconds. Per Dijkstra, computer science may be about computers as much as astronomy is about telescopes, but I would observe astronomy is pretty hard to work on without telescopes in the end.