I have led AI integration in a university faculty. From this experience I can conclude that good work is only produced when humans are in the loop. It's not a technical barrier, but a categorical one. "Good" work is defined by humans and our judgment is irrational but rooted in our evolutionary survival needs. In other words, AI don't have human motivation by definition. Without human in the loop, the top most motivation is never fully aligned with us, today, as humans. This removes the premise at the basis of this post.
Said simply: LLMs don’t have subjective reasoning.
They only “know” objective info.
Train one on your ad profile and who knows, maybe it makes decisions like you would ethics preferences and all
Agree with this. Even in software, the point of using AI is to produce something that a human finds valuable. There are many ways to use AI to build things faster, but a human has to be in the loop to point things in the right direction.
here is to you:
A) if conssumer of your service is end-user, let them write code themselves. -> result, they do not need "other human in the loop". they do not need you to develop and sell software. Replit style.
B) if consumer is AI or business, let it write it or build it themselves on demand. Codex on steroids.
C) no need to create new service at all. it all converges into single god-like super-app WeChat/Google style that does everything. eistance of different apps is history. it is all one app now.
you can very much end-up in scenario where human-in-the-loop of softwre industry is gone.
I keep thinking of the infinitely scroll GenAI app that might replace social media...
AIs are bureaucrats, no more - no no less: https://news.ycombinator.com/item?id=48321576
how many average humans do you need if your good humans are driving the ai
But that could be done by a smaller and smaller elite cadre.. maybe you could call it the philosopher Kings. But what about the masses