Home Creators Posts Import Register Favorites Logout

Downloads

Content

It's here, buds! My look at the unique visual language of art made with AI, what makes it so fascinating, and the market forces threatening to take it all away.

This is an early access version, but I think it's done. If you notice anything off, any glitches or mistakes, please let me know!

Public premiere will be tomorrow at 3PM :)

Files

Patreon version: The Surreal Dreams of AI-Generated Art

Comments

Winona Honey

I don't know if it's entirely related but this got me thinking a lot about the way the internet uses "liminal space" to describe sort of anything from a long hallway to a gas station rest stop. You mentioned the uncanny valley and I think that's similar, at least in the way that concept and liminality are talked about online. And a lot of these AI works have similar sort of almost-familiar, creepily meta-human feels to them that reminds me of being in a physically liminal space, everything is almost but not quite.

Sasha Karbachinskiy

Love this approach of taking seriously how we think about AI art and really AI production in general. It's hard for me to not see it as rapidly encroaching ubiquity, much quicker than we expect. The thing that really spooks me as that we often don't actually know what AIs are "thinking" or how they really process things. We have a unique ability to audit the stages of their "thinking" but even the programmers can't say exactly why it makes its decisions. And that's the point. Unlike coding where every aspect needs to be micromanaged by a specific command, you can let the program figure it out autonomously. It's like we've recreated the human brain and it's equally inscrutable, but private industry can and does exert power over it and over us in turn. Not to get too tinfoil-hatty about it

Henry Rachootin

The Prince idea is terrifying. The economic incentives for AI necromancy are in place, and people are excited about it.

Andrey Kurenkov

Great video! Got me to subscribe on here :) My take is that 'AI generated artworks' is a bit like saying "camera generated photographs" - AI is a tool, like a camera, and at the end of the day the art is still created by a human (only if by curation or prompting). Relatedly, your artist->programmer->photo compiler->photographers example is a little overgeneralizing, some AI artists are doing a few or all of the above themselves (see Helena Sarin - https://thegradient.pub/playing-a-game-of-ganstruction/). AI art is also deeply tied to generative/computational art, which is often overlooked - you may find this interesting https://thegradient.pub/the-past-present-and-future-of-ai-art/ . Anyway as an AI researcher I could share a lot more thoughts (eg about research orgs - they are creating tools, not art, VQGAN-CLIP being different from GANs), but all that aside again a great video, glad you covered this cool subject!