I don't want to start a flamewar, but some are suggesting we've already reached peak AI and are headed for the "trough of dis-illusionment." Since our industry seems to run on hype, what's the next technology to get on the hype-train? Seems I shouldn't start an AI startup right now since half the industry pundits say that ship has sailed.

Ideally it would be something that would sell NVidia chips (I don't want my NVDA shares losing their value) and suck down lots of power (to keep the nuclear renaissance going.)

Is there some application of fractals we've over-looked? Maybe we could have a cryptocurrency based on solving blocks of electron motility simulations so TSMC researchers could fund a "folding-at-home" style distributed system? I dunno. Just spit-balling here.