Terry Hancock (Director/Producer) on Nostr: This is pretty much my own feeling on the subject of generative AI in art. I'm not ...

This is pretty much my own feeling on the subject of generative AI in art.

I'm not entirely opposed to the concept in general. But in the specific, I have found it disappointing, mainly because it's "hard to drive".

The main source of exhaustion in doing artistic endeavors for me, is the decision fatigue. You make dozens, maybe hundreds, of little decisions. Each time you do, there's a little nudge from the "cliche shelf" of obvious and trite things to do.

Sometimes that's fine. It's just not the point, and you don't want to be original everywhere, because that gets distracting.

But when something outside of you is making one choice much easier than another, you start to slide into it. I noticed this when I was storyboarding. I'd be thinking about different possible shots, but there might be one that "would do" and was easier, because I had a photo of it. But choosing that means the photo is driving the choice, not me.

The AI brings industrial-scale automation to the cliche shelf. What was a gentle breeze becomes the backwash from a jet engine. It's pushy. It's cliched. And it is very dedicated to its cliches. It is actively difficult to get it to show a less-cliched thing. It's probably impossible to get it to do anything really original.

And unless you just fight continuously, you're going to give up and just follow the wind.

But we don't do art to do the easy thing. And we don't appreciate for that. We want to feel the artists' vision. No machine is going to give us that.

Even in some hypothetical future, where the machines are genuinely intelligent, instead of the stochastic gibberers we have now, it'll a be a machine vision, not a human one. And we humans are still going to want that human perspective.

https://www.youtube.com/watch?v=g9krnJCEBy8