Your enthusiasm about AI art reminds me of an article I just read today.
I cannot understand how, in the context of discussing AI art, you can say art is primarily about self expression. Or that you feel ownership over an AI-produced work of art. My impressions of AI art tell me the opposite is true.
A curator is NOT an artist. AI art is a mile wide and a millimeter thin. It is the visual equivalent of my personally most disliked characteristic of techno music: endless, novel combinations of elements.
Then why are you participating in a thread about AI art on a forum that is largely populated by techno enthusiasts? Surely there is a place on the Internet that caters to your preferences…
I agree with you that AI art is collaborative. But self expression, by definition, is not collaborative. Self expression cannot exist without some degree of alienation, indifference or non-conformity by the artist toward their surrounding culture. The AI artist (a computer), on the other hand, relies on constant affirmation from the outside world in the form of endless, preexisting works of art to scan. The result is impressive, but it is not expressive.
That’s what the techno enthusiasts from the social media companies believe. A future where we are only fed and feed into conversations where we are surrounded by like minded individuals.
I’m not implying that the AI is expressing itself though, but that a human is expressing themselves through it.
I find this hard to debate personally, especially when you watch the firehose feeds of those interacting with the AIs. People aren’t just asking for a picture of a nice dog, they’re conjuring up new realities with the aid of an AI renderer. And even if they did just ask for a nice dog, the AI didn’t do that, it just rendered it for them.
Think of your average Mad Men style art director. They’re never putting pen to paper, they’re directing a team - but nobody ever claims they’re not expressing themselves through a medium and they win awards for the work that they do. Saatchi and Saatchi don’t sit down at a computer with a mouse.
There are even commercial artists that have studios full of helpers doing a lot of execution and duplication work. At that point does it stop being the artists art?
If an artist loses their motor functions and then uses an AI to help them realise their visions will we really claim that’s not art? I certainly wouldn’t.
So when I as a human write a long detailed description with specific art direction, you’re willing to take away all of my agency from that situation and claim that it’s ideationless?
How?
Here’s something someone was working on yesterday in midjourney. It went through several rounds of iteration and development until they were happy with it.
How is this not creativity on the part of the one interacting with the AI? Would you claim it lacks any sense of an idea? Just random?
A giant twinky filled with chilli, in a cinematic style and soft lighting
Timothy Von Rueden nailed it down in the most precise manner and I totally aggree with him: It is not art, more of a referencing methodology to generate images based on gigantic datasets of model libraries.
Sure, one could claim that any human artist is also doing nothing else than referencing or remixing a copy of a copy of a copy of […], but this is a very cynical view on art. It undermines the personality, the experiences and the history of an artist, which is definitely shaping the results one presents as art.
It is brilliant for concept designs, quick inspirations, sketching, prototyping, as a blueprint, for story- / moodboards etc., and I would agree on calling the human part of it art direction, but my strong opinion is that neither art directors, nor creative directors or curators are artists. Sure, they are creative conaisseurs, but they don’t have to deal with the hands on part of making art.
Therefore I would rather call it a reference, or AI generated images.
I also work a lot with generative design techniques for my audiovisual works, but then again there is a process of programming, deciding, evaluating, optimizing, personalizing and kind of mis-en-scéne aspect to the works, which is out of my control during generating AI images based on prompts. So comparing these two processes is also not really making any sense to me.
Long story cut short: to call AI generated media art is a very cynical approach to art itself, that’s my personal opinion
Isn’t that just another way to describe a data set?
This is getting quite philosophical.
That’s exactly how more involved AI art direction works - and somebody prgrammed the AI. What you’re describing sounds almost indistinguishable to me, at least in a meaningful way.
I think it’s clear there are fundamental dissagrements between us about what constitutes a creative process, so I suspect a big part of this debate will simply be agreeing to disagree
By the way I think these are some good points, this is also where I see the power - we’re getting into it here and maybe your’re right that it’s as much about needing new language to describe these relationships - this may be a situation in which people debating align fairly closely in principle but just sound like they’re coming at things from a different perspective.
I work in design and the whole ‘art vs design’ debate has come up enough in my life that if I get so much as a whiff of it that I leave the room.
I aggree that this is a philosophical question, but also a question of stance to that matter: Seeing life itself as a cybernetic regulatory machine system sounds too simple to me, therefore I personally would never compare the complexity of experiences and emotions to raw data. I am, by far, not religious, but I am aware of the complexity of the mind and don’t aggree on cybernetic or transhuman technophilia, which constantly tries to degrade life to pure causality chains.
Also there is the most important part of determinism: A machine cannot do anything without an initial human input. It can evolve, regulate itself, generate instances, optimise etc., but it cannot do any of that without a clear set of algorithms and rules. Therefore one could maybe claim that the programmers of AI models are the real artists at the end, but anyone who uses their AI would, in my opinion, be some kind of assistant.
I think this is also why AI art is (at this very moment) always on the uncanny side: It cannot generate a logical form of body parts since it does not understand codes or emotions: It just refers to datasets.
Eric Wayne perfectly describes this problem by his analysis on the “insect man” he drew, compared to the result of midjourney.
It’s interesting I think you could make a convincing argument that both the AI and the human in that process are the assistant - I feel like it could depend on the context.
I could see an artist being blocked. Let’s say a concept artist. They’ve already got some kind of constraints to adhere to, or requests to fill…they could plug those into some AI and possibly get some inspiration for some approaches.
did not see this channel yet, I will look into it, thanks!