This seems weird to me. Isn’t Japanese culture obsessed with aesthetics? Like when you buy fast food there it actually looks neat and clean like in advertisements, and you might spend a small fortune on a couple of strawberries but they’ll be the brightest, most flawless specimens you’ll see in your lifetime. Even cheap throwaway meals are packaged like a professional food stylist went over them.
That focus on perfection seems incongruent with the slop machine that churns out soulless, derivative, and often obviously flawed ‘art’.
There’s a big difference between what’s happening in the commercial world and art circles. I can’t speak for food but with regards to the topic here, Japan has a huge creator market and that market responds very differently. A year back, their biggest creator fair had a few “artists” selling AI “art” but essentially no one was interested and they hardly sold anything while others had sprawling queues. This year there was pretty much 0 as a result.
Maybe they have high standards for what they generate, use the models as “fillers” or “finishers” (broadly adding details or patterns to a base sketch, then touching up on any errors or ungeneratable details), or rely on models they’ve vetted for consistent quality? I doubt they’re all using a piss filter chatgpt text to image gen.
Japanese creatives are now heavily split on LLMs. What with half taken to defend their work with watermarks and proofs of timelapses and manual processes, and there’s fierce loyalty for Miyazaki et. al. While others are either playing with dozens of LLMs to produce even animations, churning out thirst traps by the dozen, and still others “enhancing” game screenshots of their player characters in poses and skimpy clothing.
This seems weird to me. Isn’t Japanese culture obsessed with aesthetics? Like when you buy fast food there it actually looks neat and clean like in advertisements, and you might spend a small fortune on a couple of strawberries but they’ll be the brightest, most flawless specimens you’ll see in your lifetime. Even cheap throwaway meals are packaged like a professional food stylist went over them.
That focus on perfection seems incongruent with the slop machine that churns out soulless, derivative, and often obviously flawed ‘art’.
There’s a big difference between what’s happening in the commercial world and art circles. I can’t speak for food but with regards to the topic here, Japan has a huge creator market and that market responds very differently. A year back, their biggest creator fair had a few “artists” selling AI “art” but essentially no one was interested and they hardly sold anything while others had sprawling queues. This year there was pretty much 0 as a result.
Maybe they have high standards for what they generate, use the models as “fillers” or “finishers” (broadly adding details or patterns to a base sketch, then touching up on any errors or ungeneratable details), or rely on models they’ve vetted for consistent quality? I doubt they’re all using a piss filter chatgpt text to image gen.
Japanese creatives are now heavily split on LLMs. What with half taken to defend their work with watermarks and proofs of timelapses and manual processes, and there’s fierce loyalty for Miyazaki et. al. While others are either playing with dozens of LLMs to produce even animations, churning out thirst traps by the dozen, and still others “enhancing” game screenshots of their player characters in poses and skimpy clothing.
Have you considered that wherever you’re from is the weird place? They are much more optimistic about AI in Asia.