Hacker News Reader

Comment by whateveracct | Hacker News Reader

This feels like a comment written by someone without practical experience with concept art (creator or consumer) - correct me if I'm wrong though!

Like personally, I think I will still draw concept art instead of asking a GAN. Feels like that greases my creative wheels more. Just me though.

View on HN

You are wrong, I have worked for many years as an artists and level builder in video games. Bioshock is the most famous game, and Void Bastards the most recent.

Could you instead use AI to reduce drudgery in your work, or to iterate faster between ideas until you settle on a final peace? What if you have something in mind that is hard to prompt for, or the tool just refuses to generate what you had in mind? I would try to lean into it as a tool to speed up some of the tedium, instead of fear it, if possible

I'm not in the camp that fears it, I think it will be an awesome tool and I can't wait to use it in my next game. I've been playing with Midjourny a lot already, but have been watching Stable Diffusion and the control net stuff closely as well.

Whether you have experience or not, game companies already use AI for concept art.

not only art, Obsidian is using already for voices[0]. The future is here.

0:https://www.youtube.com/watch?v=YajBa5PO1Hk

I could see drawing a couple of characters and getting an AI to extend the style to more. Especially if you can give it feedback. But the idea that you could get years of concepts in an afternoon is silly. That's like saying I can generate hundreds of business plans using ChatGPT.

Reply on news.ycombinator.com