5.03.25

why is this post so long? i've been thinking about how the easiest form of getting New information these days is through micro blogging (twitter/bluesky). while i appreciate the 240 character limit a lot, i realized i haven't visited the library in a while, and this was a good opportunity to do it!

major disclaimer!

this blog post is about the use of generative AI in art (primarily audio and visual). this blog does not concern its environmental impacts, copyright issues, or deeply irresponsible and dangerous usage in misinformation, as I feel these have been covered enough already. nor does this blog concern the use of AI in other fields (...i think it's pretty agreeable to have computers classifying traumatic content instead of humans). in reality, i believe the push for generative AI has serious consequences for the livelihood of human artists and can destabilize the standards and expectations of our audiences. i could write as much as i want about the artistic and philosophical issues of AI but none of it would matter if companies suddenly decided to fire all of their artists and no one cared and also the Earth died.

i think the best way to describe this blog would be a meditative attempt to understand why i feel so uncomfortable about the use of AI in art without being entirely reactionary... this dive has been much deeper than expected for me. i've changed my "thesis" multiple times since i started thinking about this in February and i think it's bound to change again. i may update this blog post or make an entirely new follow-up.

also, i'm not a philosopher or a computer scientist or anythingggg i just think about this stuff because in the past year ive seen my final semester project partner use AI to finish our paper (reads bad) and ive seen music producer friends use AI for their album covers (looks bad) and ive walked into my boss's office to see them using AI to convert inches to centimeters (?????) and i dont know whyyy!!!!!!!!!!

if you are looking for some relief in the midst of the AI boom, my immediate, actionable 2-cents is this: the upside of living in an information-rich age is that although it costs you very little to Generate whatever you want (for now!), it also costs you very little to enjoy real art with direction and intent.

but in our current situation of decreasing computer literacy and increasing anxiety about time (these two issues are related), perhaps the real issue at hand is finding art. not just the act of finding it through search engine slop, but wanting to find it. wanting to look beyond what you think you want, to look beyond yourself. wanting to see a work that might not immediately serve you, and being okay with that. one of art’s greatest powers is the way it transforms you in ways you may not expect—and when generative AI is inherently designed to give you exactly what you want to see and what you want to hear and what you want to read, what exactly are you gaining?

i still believe there are people out there who still look for that transformative power. i also believe that there are even more people who are capable of thinking this way and just don't realize it! in a world that desperately clings to nostalgia in order to justify inaction and stagnation, i implore you to channel that nostalgia in a way that frees you. (in my case, it was making a personal site and more music!)

in the pursuit of trying my best to come up with a futureproof, non-contextual answer for why i find it so uncomfortable that people reach for AI so easily and so genuinely, i realized that futures and context might have been what i was actually so concerned with.

for the purpose of this blog post, AI-generated art is art. PAUSE!!!!!!!!!!!! this doesn’t mean that i think AI-generated art is “good” art or immediately worthy of positive response… only that it's subject to the same kinds of critique and virality as the other forms of art that came before it. this inclusion is especially relevant in our current age where multiple kinds of art coexist under the banner of "content" for social media feeds.

by conceding serious ground, this interpretation also allows room to look at events that have already happened, like the generative AI ash baby being used to convey human reactions and people generally finding Mr. Chedda funny. though it's recognizably shiny garbage 99.99% of the time upon closer inspection, generative AI images, and generative AI art by extension, are absolutely capable of being indistinguishable from human-made things and thus capable of eliciting the same kind of response. in this way it becomes more difficult to justify the idea that machine-generated images inherently lack "soul", or an intrinsic quality that resonates with all humans... and the inverse idea that human art inherently has such a quality. this is what the rest of this blog post grapples with.

replacement and future

so why do i still feel so uncomfortable about this supposed “lack of soul”? is it the fact it's not "real humans" doing "real art"?

the progression of digital tools do often come with cruel replacements. replacing entire musical ensembles with a kontakt library sounds very cruel. but in practice, it’s not controversial; unless they're hans zimmer, no one expects a bedroom producer to have an entire ensemble of human musicians at their convenience. replacements can even become artistic statements of their own. 30 years later, there's still something very fresh and genuine about the sound of the Yamaha DX7 and the Korg M1 despite being “simulations” of real instruments.

replacements don't always lead to extinction. high quality photoshop jobs didn’t replace photography, and high quality photography didn’t replace painting. bands still record and perform using real instruments, and the synthesizer became another instrument in itself. new forms and layers of expression don't need to detract from what was there before. it's as English artist Brian Eno says: “whatever you now find weird, ugly, uncomfortable and nasty about a new medium will surely become its signature.” as soon as new mediums are replaced, they find a home somewhere.

my prediction, or maybe already an observation: generative AI art isn't an exception to this. but once the AI boom dies down like it always does, the definitive home of generative AI will be in the fucking trenches of the Internet. it's already being used to get thousands of likes from senile Facebook users in the same way that JPEG-obliterated boomer comics do. i've seen generative AI used for:

• "how it feels" / "me when" reaction images
• reinvigorating "that friday feeling" wine mom memes
• background images of Youtube mixes and song re-uploads
• "aesthetic" pages on social media that cite Pinterest as credit
• ragebait

generative AI is just a utility for making the lowest common denominator forms of content on the internet. things that already exist. the fact that i would really not care if these listed things disappeared, human-made or not, is probably the most subjective part of this blog and if you don't agree then i don't really know how to defend myself...

this is the near future of generative AI. it wouldn't be surprising if, a decade from now, the general online populace is nostalgic for it in the same way i find comfort in mislabeled trance videos from 2010 with uncredited anime girl art as the background. but maybe generation alpha will be fond of generative AI for the reason we generally aren't now: the "weird, ugly, uncomfortable and nasty" aberrations that distinguished it most clearly in its nascent stages. they’ll be nostalgic for the extra appendages, the hallucinations, the "dreamlike" haziness... none of the preexisting human art or photography it attempts to emulate. but for now, the push by developers to eliminate these aberrations in newer models weakens the legitimacy of generative AI as an artistic medium rather than as a redundant utility.

authentication by taste (audience side: still a search for the soul)

indeed, for now, if the broad and profitable goal of generative AI art is to fully emulate preexisting human art and photography—if its developers so badly want it to be respected in the same regard—then it can be analyzed and criticized in the same way. not only can it be evaluated by its contents, but it can also be stripped apart for its intent and context. and when the original intent of generative AI artwork is most often to copy something that already exists without any additional insight, it wouldn’t be unreasonable to dismiss it when the original work could be studied instead... or, spiritual successors that synthesize new takes on the work from their own experiences or tastes. just as with human-made art, if it doesn’t offer anything different, it’s not likely to stick to the heart (see: movie reboots and video essays).

context is also part of this investigation, i.e., an emotional or social authentication of the work. an album can have social authentication if it is recommended by a favorite artist or a friend (or made by one ^_^). more generally, your favorite long-standing curators of art (small labels, blogs) can likely still be trusted. a drawing or a photo that may not be immediately self-explanatory can have emotional authentication if its backstory is compelling. what kinds of people recommend generative AI art or music to their friends? has anyone ever been moved by the story behind a generative AI piece? (... we’ll see how the answers to these questions change in the next few years...)

perhaps, after all, the search of intent and context is the search for “soul”... but it’s a deeply personal search, one that can’t be applied evenly (should Feel It Motherfuckers and Tails Gets Trolled be evaluated on the same level? i really love both.. and i don’t think gen AI could make either one). such a subjective lens that allows both artists and audiences to form their own tastes in art is also valid in the fight against generative AI art.

trusting tools (artist side: still a search for the soul)

trust is also an integral aspect in the creation of art. more concretely, artists choose who and what they work with. they choose who to feature, what to draw with, who to sample, what instrument to play. this kind of trust doesn’t need to be so serious; it can be a matter of convenience or comfort, like picking up a guitar just because you have it. but consciously or not, by inviting someone or something else into their vision, the artist shows that they trust or value the input. it’s a more internal form of authentication, specific to the artist. if generative AI is an artist, a medium, a collaborator in the place of a human, then

• do you personally accept its level of quality,
• do you respect it and admire it,
• or is its interpretation at least special or personal to you in some way?
• do you trust generative AI?

i wrote these questions out in the hopes that both artists and audiences can remember that we aren’t held entirely hostage to convenience or algorithms. other than that, all i have come to find out is that the fight against generative AI in art is the same fight for the soul, something so fleeting and fragile that it is only up to you to know what it looks like...

my own personal answer: i don't trust generative AI to have as much fun as meeeeee!!!!!!!!!!!!

♥ 14.4

a list of things i read / watched while writing this blog post, like an informal citation page.. please don't throw tomatoes at me lol

1. anne hero blog 012: psychic armor for the modern girl. if you're looking for something more actionable and what to do to take more control of your experiences with the Computer, this is a great start!!!
2. Reality+: Virtual Worlds and the Problems of Philosophy. by David J Chalmers. i'm actually still working through this book, and as a very Do Things person, some of it is definitely going over my head... but this book's section on deepfakes introduced to me the idea of "authentication" for digital media.
3. You are a better writer than AI. (Yes, you.). a video recommended to me by a friend... thank you :) it's about the generative AI and human connection from a writer's perspective.
4. Petscop's Spiritual Sequel: 3D Workers Island (FULL ANALYSIS). this doesn't have too much to do with the blog post but i liked 3dwiscr and clicked on it. the AI and Consciousness section is maybe one of the first things that got me thinking about how we now exist in a feedback loop with our computer Things, making it much harder to distinguish the human-made from computer-made (is there really such a difference?).
5. Mark Fisher : The Slow Cancellation Of The Future. (RIP). one of the things about generative AI art that annoys the hell out of me is art of something that already exists in abundance, like 1990s - 2000s anime art. The generation of images that could easily be stock or clip art, but people insist on using the gunk machine to produce gunk that is as-close-to-non-gunk-as-possible (ridding of the abberations). i think this could be related to how Mark Fisher talks about an insufferable and perfect perpetuation of the past as a shield from a really New future. i'm not exempt from the phenomenon (i make happy hardcore...), but i do hope it holds more weight in the context of generative AI.

...and more probably that i haven't remembered yet...