On the other hand, as both an artist and machine learning practitioner, I think most artists are only seeing the surface layer here and have little insight on its derivative, the algorithms and state of research which are advancing the state of the art on a weekly basis. It'll never be obvious that we're at the critical moment because critical phase changes happen all at once, suddenly, out of nowhere.

There is an insane amount of low-hanging fruit right now, and potentially decades or centuries of very important math to be worked out around optimal learning strategies, but it's clear that we do have a very strong likelihood of our ways of life being fundamentally altered by these technologies.

I mean already, artists are suddenly having to grip with all sorts of new and forgotten questions around artistic identity and integrity, what qualifies as art, who qualifies as an artist... Generative technology has already made artists begin to question and radically redefine the nature of art, and if it's good enough to do that, I think it's already worth serious consideration. These technologies, even in current form, were considered science fiction or literal magic up until very recently.

> I think most artists are only seeing the surface layer here and have little insight on its derivative (…)

> Generative technology has already made artists begin to question and radically redefine the nature of art (…)

So which is it? Are artist not understanding the potential of the technology, or are they so flabbergasted they are redefining the nature of art? It can’t be both.

> There is an insane amount of low-hanging fruit right now

Precisely. What I’m criticising is the generic low-effort response of assuming that being able to pick low-hanging fruit now indicates with certainty that the high-hanging fruit will be picked soon. It doesn’t. As an exaggerated analogy, building the first paper airplane or boat might’ve been fun and impressive, but it was in no way an indication we’d be able to construct rockets or submarines. We eventually did, but it took a very long time and completely different technology.

To really drive the point home, my comment wasn’t specifically about art or LLMs or the user I replied to. What I am against is the lazy hand-wavy extrapolation which is used to justify anything. As long as you say it’s five years away, you can shut down any interesting discussion.

> These technologies, even in current form, were considered science fiction or literal magic up until very recently.

I don’t recall science fiction or magical stories—not any that weren’t written as cautionary tales or about revolution, anyway—which had talking robots which were wrong most of the time yet spoke authoritatively; convinced people to inadvertently poison or kill themselves and others; were used for spam and disinformation on a large scale; and accelerated the concentration of power for the very few at the top. Not every science fiction is good. In fact, plenty of it is very very bad. There’s a reason the Torment Nexus is a meme.

> Generative technology has already made artists begin to question and radically redefine the nature of art,

Everyone else still knows what art is. The only reason artists grapple with it is because it's existential for them. Artists think of themselves as the gatekeepers of art but the only thing that qualified them for that (in their minds) was the ability to produce it.

Now that everyone is producing generic entry level art (what most artists do anyway), they are losing their identity. The "What is art?" conversation isn't about art, it's about gatekeeping.