Facing The Enemy.
Added 2026-02-28 13:46:39 +0000 UTC
Comments
I'd say with the experiment concluded: Rendering of the characters is right, but as you pointed out it gets the details wrong: Eye alignment fails on the page itself and usually with no expression. Reggie just seems to alternate between Lupin III and half elf, which weirded me out the most. Anyways it still doesn't hold a candle to the real deal. But I also wonder for how long...
TiredGerman
2026-03-02 10:08:33 +0000 UTCI used AI recently and I see its potential.
Went Brown
2026-03-02 02:22:18 +0000 UTCAI always seems to get just enough wrong to stay in the Uncanny Valley for me, luckily ... with these (and often elsewhere too) it's the eyes, they seem out of alignment? Esp. for multcharacter pics - the 1000 mile stare is past the other character. Worth an experiment, but posed art won't replace storyline Jackie!!
cee
2026-03-01 18:39:40 +0000 UTCGenerative AI is in the uncanny valley side of things, where it seems fine, but there's something off, something weird. A strange feeling you can't shake of... you used your own pieces on Grok and I'll be honest, almost None of the comic pages turned out better than the source. They feel soul-less and and very goofy (Oh no! they whitewashed Jo!). with the exception of those that seem like upscaling your work. It's a whole discussion we are not having! AI canot replace that tried and true Jackie wit and life experience
NeoChomik
2026-03-01 18:35:23 +0000 UTCI def don't aupport AI, but.... Some mixed body-type issues with Carol and Alex in there, interstingly. Jo and Jess seem to "translate" best, which is fine by me, Jo being my personal favorite...
Ascender
2026-03-01 16:59:36 +0000 UTCI think the best use for AI in my case is making it show me how to set up lighting and things that I have trouble visualizing.
Jackie Wohlenhaus
2026-03-01 09:03:42 +0000 UTCI could probably edit these to fix the issues, but it's not worth is since basically no one has expressed any desire for this.
Jackie Wohlenhaus
2026-03-01 09:01:56 +0000 UTCI'm sure that being able to spot AI images is some kind of survival warning.
Jackie Wohlenhaus
2026-03-01 08:57:07 +0000 UTCYou and me both, bro.
Jackie Wohlenhaus
2026-03-01 08:55:19 +0000 UTCI'm doing a book cover for a client. On a whim, I showed it to an ai and told it to do it better. It did it better. I am... I think discouraged is the right word.
William Caddell
2026-03-01 05:29:03 +0000 UTCAlex looks more attractive here than in the comic, which is something I never thought I'd say. But there is something...empty...in these, as others have noted.
ValdVin
2026-03-01 01:35:03 +0000 UTCI'm surprised you can read well enough to reply if this is your response.
Jackie Wohlenhaus
2026-02-28 22:39:34 +0000 UTCAriamaki
2026-02-28 22:34:58 +0000 UTCI'm enjoying seeing the AI versions of what you've already given us, because I've already seen your original version. But I'd get really tired really quick if I were reading just the AI version. There's just something a little... off. I think the term is "uncanny valley". The AI version is supposed to look real, but it misses the mark. But it's too close for the brain to reconcile the image and make it fully real. It's like those filters people use on snapchat and whatever to make their faces look unnatural. Even though I know what the real person looks like, my brain can't seem to fix their unnatural photo. But with the AI version of what you've already given us, my brain can remember what your original drawing looked like, and I can translate that to the "real" picture that my brain created from your artwork.
awgiedawgie
2026-02-28 17:11:15 +0000 UTCI've done the same thing with some of my own art, to see how it does creating a 3D model that I can manipulate to get better results. But you always have to go in and fix things to get expressions. Sort of like the comic-maker software where you create character models.
Paul Osze
2026-02-28 16:09:31 +0000 UTC
