I'm not sure why we have to spend so much time explaining and justifying using AI tools in art creation. Let's just experiment, play, and enjoy it. Whether the gatekeepers deem it "art" or "slop" is entirely up to them and not our concern.
I like this conception, but I also think there's a spectrum between curation and creation that is not appreciated in the discourse about this subject, perhaps partly because of the UX of generators most people have encountered. The UX of ChatGPT Images is that you ask the AI to make you an image, maybe based on one or more images you supply, giving as much or as little detail as needed, and then in an ideal world you get out the image you wanted, or even something better than what you had in mind. You can ask the AI to make changes, either in the chat or by making comments on parts of the image, but it's nonetheless a pretty hands-off process.
There is a whole other world of stuff you can do if you have more control over the diffusion process with local models or a more expansive API than the one OpenAI provides. Setting the denoise level and targeting specific parts of the image you're working on lets you do everything from smoothing out lines to completely altering the style and/or content, and there are tonnes of other techniques like ControlNet (famously responsible for Spiral Town), regional prompting, edit models and so on. There's a plugin for the Krita painting program that I think represents the best implementation of this idea, and plenty of YouTube videos of people using it to visually shape the AI's outputs in real time.
I think even sufficiently detailed prompting probably reaches a point where it's more than curation. Ideogram recently released an open weight model that's trained on a specific JSON format for prompting which allows you to specify bounding boxes for elements of the generation, and you can get very granular with it and really control the composition. https://ideogram.ai/blog/ideogram-4.0/
I'm hesitant about advocating some kind of Labour Theory of Value for art where the amount of human effort required to make it determines the work's value, but I'm really interested how works produced with the tech can blur the line between curation and creation like this. Personally, I lean towards wanting to create and get the output as close to my initial vision as possible, but an undeniable part of the fun of AI image generation is when the model surprises you with an unexpected interpretation of your prompt that may be better than what you first had in mind.
Agreed. I do my AI art locally in ComfyUI, and the amount of complexity and effort needed to get something specific is easily enough to qualify it as "real" art, even with the moved goalpost definition that some people use nowadays.
I wasn’t sure if I was going to agree with this at first but your “neat rock” analogy feels very accurate to me, and you’re absolutely right about Pitchfork. Also several of those images you posted absolutely made me feel something. They felt familiar and nostalgic for something in my own life.
I think a lot of people’s exposure to AI images is still the very sloppy stuff and they don’t know how specific the images can get, it can be pretty wild to play around with them.
I really value this perspective. As with most things AI, we are moving to meta-cognitive activities that are farther reaching and more expansive in how we engage in information, judgment, and - as you out here - taste and cultivation.
And, as the weirdo who unironically likes contemporary and conceptual art: curation, randomness, collage, found objects, even the deliberate rejection of creative control and intent, even _literally doing nothing_ - they're all perfectly valid as art. Every supposedly mandatory element of art has been rejected by someone, somewhere, whose work is now in a museum (to be scoffed at by someone else, saying "my kid could do that").
I think Suno is particularly fascinating. With an AI image there's often some debate about how much "you" went into it, whether your prior intent was realized (you actually "created" it... but you didn't "make" it... maybe?) or whether you just loved the result (you "merely" curated it!). With Suno, you have the full range of "uploading one track of my own demo tape and my own lyrics to enhance the sound", to "tweaking and refining Suno's song and ChatGPT's lyrics and remixing the tracks until it makes me maximally happy", to "curating variations of guitar riffs", to "playing the slot machine for the lols".
And nobody but you can tell which it was.
We'll just have to get used to what people in contemporary art have known for decades; art can include every permutation of human involvement, including those where it comes only at the very start of the process (pure ideation as art) or only at the very end (curation as art).
I would argue that creativity is only one part of art and creation and curation; there are other skills needed, too, such as aesthetics, artist skills, vision, etc.
Is a skilled session musician who never writes his own songs creative? Is a poor musician who writes 100s of songs creative?
Perhaps the word “creativity” is very open to interpretation. I used to work in a “creative” job: photographer in an advertising agency, but the “creative” part of the job was mainly copying others. Now, as a teacher, I have to be much more creative every day, but I doubt many people would mention this career if they had to list ten creative jobs.
Looking at your AI creations (note the word “creation”), it is clear that they contain the skills I mentioned above; even if they are all rolled into excellent prompting, you have to have the vision and the artistic understanding to be able to know what you want (AI) to create.
The analogy I've always used is that we've suddenly gained access to bring back artifacts from the Platonic realm of all possible arrangement of unicode-characters/pixels/frequencies/whatever, and so we're in a sort of gold-rush archaeology era.
We tell our little robot friend what sort of pixels (or unicode characters or sound frequencies or whatever) to look for, it goes into the Platonic realm and looks for them, and brings back what it finds. And if we think it's cool enough, we show what we found to all our friends.
And it really does feel like archaeology to me, like I'm getting glimpses into an alien culture that exists in an alternate branch of the multiverse that we can only partially reconstruct the details of here. Like Suno will produce an anime opening theme song, and I'll really wish I could have watched that counterfactual anime, it seems like it would have been an awesome show.
I am happy to see that you are broadening your repertoire of bad takes about AI beyond the environmental concerns.
BTW, on those environmental concerns, the source that you used for the numbers on that calculator has a major error in it. It assumes that Claude is a MoE, and rates it as the most energy efficient of the big three.
Claude is definitely not a MoE and as such is the least efficient of the big three by far.
I'm not sure why we have to spend so much time explaining and justifying using AI tools in art creation. Let's just experiment, play, and enjoy it. Whether the gatekeepers deem it "art" or "slop" is entirely up to them and not our concern.
I like this conception, but I also think there's a spectrum between curation and creation that is not appreciated in the discourse about this subject, perhaps partly because of the UX of generators most people have encountered. The UX of ChatGPT Images is that you ask the AI to make you an image, maybe based on one or more images you supply, giving as much or as little detail as needed, and then in an ideal world you get out the image you wanted, or even something better than what you had in mind. You can ask the AI to make changes, either in the chat or by making comments on parts of the image, but it's nonetheless a pretty hands-off process.
There is a whole other world of stuff you can do if you have more control over the diffusion process with local models or a more expansive API than the one OpenAI provides. Setting the denoise level and targeting specific parts of the image you're working on lets you do everything from smoothing out lines to completely altering the style and/or content, and there are tonnes of other techniques like ControlNet (famously responsible for Spiral Town), regional prompting, edit models and so on. There's a plugin for the Krita painting program that I think represents the best implementation of this idea, and plenty of YouTube videos of people using it to visually shape the AI's outputs in real time.
I think even sufficiently detailed prompting probably reaches a point where it's more than curation. Ideogram recently released an open weight model that's trained on a specific JSON format for prompting which allows you to specify bounding boxes for elements of the generation, and you can get very granular with it and really control the composition. https://ideogram.ai/blog/ideogram-4.0/
I'm hesitant about advocating some kind of Labour Theory of Value for art where the amount of human effort required to make it determines the work's value, but I'm really interested how works produced with the tech can blur the line between curation and creation like this. Personally, I lean towards wanting to create and get the output as close to my initial vision as possible, but an undeniable part of the fun of AI image generation is when the model surprises you with an unexpected interpretation of your prompt that may be better than what you first had in mind.
Agreed. I do my AI art locally in ComfyUI, and the amount of complexity and effort needed to get something specific is easily enough to qualify it as "real" art, even with the moved goalpost definition that some people use nowadays.
I wasn’t sure if I was going to agree with this at first but your “neat rock” analogy feels very accurate to me, and you’re absolutely right about Pitchfork. Also several of those images you posted absolutely made me feel something. They felt familiar and nostalgic for something in my own life.
I think a lot of people’s exposure to AI images is still the very sloppy stuff and they don’t know how specific the images can get, it can be pretty wild to play around with them.
I really value this perspective. As with most things AI, we are moving to meta-cognitive activities that are farther reaching and more expansive in how we engage in information, judgment, and - as you out here - taste and cultivation.
Couldn't agree more.
And, as the weirdo who unironically likes contemporary and conceptual art: curation, randomness, collage, found objects, even the deliberate rejection of creative control and intent, even _literally doing nothing_ - they're all perfectly valid as art. Every supposedly mandatory element of art has been rejected by someone, somewhere, whose work is now in a museum (to be scoffed at by someone else, saying "my kid could do that").
I think Suno is particularly fascinating. With an AI image there's often some debate about how much "you" went into it, whether your prior intent was realized (you actually "created" it... but you didn't "make" it... maybe?) or whether you just loved the result (you "merely" curated it!). With Suno, you have the full range of "uploading one track of my own demo tape and my own lyrics to enhance the sound", to "tweaking and refining Suno's song and ChatGPT's lyrics and remixing the tracks until it makes me maximally happy", to "curating variations of guitar riffs", to "playing the slot machine for the lols".
And nobody but you can tell which it was.
We'll just have to get used to what people in contemporary art have known for decades; art can include every permutation of human involvement, including those where it comes only at the very start of the process (pure ideation as art) or only at the very end (curation as art).
I would argue that creativity is only one part of art and creation and curation; there are other skills needed, too, such as aesthetics, artist skills, vision, etc.
Is a skilled session musician who never writes his own songs creative? Is a poor musician who writes 100s of songs creative?
Perhaps the word “creativity” is very open to interpretation. I used to work in a “creative” job: photographer in an advertising agency, but the “creative” part of the job was mainly copying others. Now, as a teacher, I have to be much more creative every day, but I doubt many people would mention this career if they had to list ten creative jobs.
Looking at your AI creations (note the word “creation”), it is clear that they contain the skills I mentioned above; even if they are all rolled into excellent prompting, you have to have the vision and the artistic understanding to be able to know what you want (AI) to create.
The analogy I've always used is that we've suddenly gained access to bring back artifacts from the Platonic realm of all possible arrangement of unicode-characters/pixels/frequencies/whatever, and so we're in a sort of gold-rush archaeology era.
We tell our little robot friend what sort of pixels (or unicode characters or sound frequencies or whatever) to look for, it goes into the Platonic realm and looks for them, and brings back what it finds. And if we think it's cool enough, we show what we found to all our friends.
And it really does feel like archaeology to me, like I'm getting glimpses into an alien culture that exists in an alternate branch of the multiverse that we can only partially reconstruct the details of here. Like Suno will produce an anime opening theme song, and I'll really wish I could have watched that counterfactual anime, it seems like it would have been an awesome show.
I am happy to see that you are broadening your repertoire of bad takes about AI beyond the environmental concerns.
BTW, on those environmental concerns, the source that you used for the numbers on that calculator has a major error in it. It assumes that Claude is a MoE, and rates it as the most energy efficient of the big three.
Claude is definitely not a MoE and as such is the least efficient of the big three by far.
If listening to rock n roll with your friends while enjoying beer and barbecue is wrong, then I don’t want to be right
We’re all curators aren’t we. Letterboxd, goodreads, Spotify playlists. We all want to share our taste and usher you over to our table.
What’s the giveaway there?
Against some rival /mu/ factions definitely
... are you a bot? What is this reply?