Software is becoming less and less of a tech industry and more and more commoditized. Very little innovation in the past 5-10 years. Very few successful companies during that time have come about by having incredibly innovative and novel technology - almost all are simply using run of the mill, commoditized technology to solve business problems that they have large domain knowledge of. Given this, it makes total sens…
We're moving up on the abstraction ladder, and as hacker news has always had an early adopter audience, focus is shifting here first.
Assuming these trends are highly correlated with the Kondratieff cycle, I expect a new trend to emerge about a decade from now. My best bet is focus will shift deep into the implementation again, but now in another context (Maybe on edge devices, data/code floating in the void, more decentralized, ...).
I've put my money where my mouth is, and my personal bet is on interactive 3D, as 3D capable devices are becoming more of a commodity and the current generation of youth (future buyers) will be a very technically literate audience that has grown up playing games and being connected 24/7.
Maybe I should send myself a reminder about this post in 10 years from now; curious to see what will have emerged by then on this site...