What AI art can't do. And why maybe that doesn't really matter :-(

[zakael19]

This reminds me of the dangers of temp-tracking during film production. This is when the director, editor, et al will drop in pre-existing music while editing a movie. They do this to get feel for what the film or a scene will feel like, and/or to suggest to a film score composer the kind of music they're looking for. The problem with this approach is that a composer can't help but be influenced by the pre-existing music, especially if they're pressed for time. Over time, as a result, you wind up with movies of similar types that have very similar music in similar kinds of scenes, and it all feels very formulaic.

Ah, you too have seen the Every Frame a Painting video on Marvel eh?

I linked that with this exact idea back in one of the other AI threads here. I think that's a real risk, instead of working with an Art Director or actual artist to establish a house/product style first and then iterate on that.

 

[aramis erak]

Can I ask why temporary AI art would be worse than one with no art? What does that use signal to you, that would push you away?

It shows a willingness to "do anything" to get the funding, (do note the scare quotes,) and provides a temptation to go cheap on the final. A temptation I don't want to feed.

 

[Bill Zebub]

It shows a willingness to "do anything" to get the funding, (do note the scare quotes,) and provides a temptation to go cheap on the final. A temptation I don't want to feed.

Follow-up question (for anybody who feels like answering): would you simply pass, or would you take steps to oppose/undermine such a campaign? Even if it just means a post "warning" other people to stay away.

(If you're wondering: no, I'm not in any way planning to start a crowdfunding campaign, or even to publish anything for free. I'm just curious about the mindsets.)

 

[Artamo]

Question for the (strongly anti-AI) community: would you ever back a Kickstarter that promised to deliver only human-made art, but needed the Kickstarter to pay for that art, and therefore used AI to generate example/conceptual art, to make the Kickstarter more visually appealing?

I think there's a secondary use, as well: times when I've hired artists (illustrators, graphic designers, and also architects) I have struggled to put my requests and ideas into the correct words, and there has been a lot of (expensive) back-and-forth as I say, "No, no, that's not what I meant...". I'd rather do that iteration myself, working with AI, and then look at the results with the artist in order to have a more concrete discussion.

Personally, I feel that would actually be a significant red flag against the project, showing that good preliminary work has not been done to actually sell the concept prior to the launch. It indicates this is someone without a proven track record or has not contracted an art team to create samples.

 

[aramis erak]

Follow-up question (for anybody who feels like answering): would you simply pass, or would you take steps to oppose/undermine such a campaign? Even if it just means a post "warning" other people to stay away.

(If you're wondering: no, I'm not in any way planning to start a crowdfunding campaign, or even to publish anything for free. I'm just curious about the mindsets.)

Mildly oppose. Point out the AI use, and leave it at that. Maybe lampoon any friends who support it.

 

[Bill Zebub]

Personally, I feel that would actually be a significant red flag against the project, showing that good preliminary work has not been done to actually sell the concept prior to the launch. It indicates this is someone without a proven track record or has not contracted an art team to create samples.

I mean, that makes sense to me, too, if we're talking about something from Free League or Arcane Library, where it's aiming to be another million dollar campaign.

But there are countless small-scale projects that might raise a few thousand dollars. In those cases I doubt investing in preliminary work is a wise investment.

 

[Artamo]

I mean, that makes sense to me, too, if we're talking about something from Free League or Arcane Library, where it's aiming to be another million dollar campaign.

But there are countless small-scale projects that might raise a few thousand dollars. In those cases I doubt investing in preliminary work is a wise investment.

I would think that the scope of the project also would set the level of upfront work the creator is putting into it. I would expect a lot more from Free League or Arcane Library today - like a lot a lot. For something that only requires a few thousand, how much sample art could really be needed? If it’s truly something the creator wants to see get done, I would think they’re putting some skin in the game to get it launched.

 

[Morrus]

Follow-up question (for anybody who feels like answering): would you simply pass, or would you take steps to oppose/undermine such a campaign? Even if it just means a post "warning" other people to stay away.

(If you're wondering: no, I'm not in any way planning to start a crowdfunding campaign, or even to publish anything for free. I'm just curious about the mindsets.)

We wouldn't promote it on EN World or the podcast or anything.

 

[Bill Zebub]

We wouldn't promote it on EN World or the podcast or anything.

Oh, yeah, I assumed that. That's still just "not supporting" in my book, though, as opposed to "trying to undermine."

 

[Dire Bare]

Oh, yeah, I assumed that. That's still just "not supporting" in my book, though, as opposed to "trying to undermine."

If the project seems to have promise . . . but uses generative AI in any way . . . I MIGHT post a comment somewhere relevant stating exactly that, "Hey, project seems cool, but your use of gen-AI is a no-go for me. Sorry".