StillDrinking has a lot of great essays. He has a way with words that makes reading his articles both entertaining and illuminating at the same time – a powerful combination.
Anyway, their essay on AI and ChatGPT, and how the world we have built will continue manages to be human-hostile in many ways in spite of (and indeed, in part because of) technology:
Tech Erosion
People are already losing their jobs. It’s not only the artists, whom nobody cares about until they’re gone, it’s copyeditors and clerks and designers. And just like self-checkouts and airport entry surveys, the humans are replaced by something a little bit worse. But it’s cheaper, and novelty often obscures indignity long enough for it to entrench, and we all accept that everything is a little bit slower, a little bit less trustworthy, and everything has a little more friction to grind us down over each day. The replacement bots could be honed into better tools, but who will bother once they’re accepted? Market trends always converge on giving us as little as possible.
I’d also be amiss if I didn’t highlight this crucial insight into our collective fears of AIs:
The terror of building a super artificial intelligence is not due to having something super intelligent hanging around, it is the terror of having something super intelligent that acts like a human. Because if we manage to build something technologically superior to us that also acts like us, it will do what technologically superior humans always do to their neighbors.
Perhaps there are things that we, humanity as a whole, should just decide that we should not build – not because we cannot or because it would be evil, but because we are not sure it’s a net positive as a whole. Of course, if we were able to choose so, we probably would have less to be concerned about already, given that we’d clearly have a mechanism for informed, preventive, and collective action.