As an art photographer (not in a professional capacity, but professionally trained), I naturally have an opinion on this. It is, after all, one of the few fields in which I actually somewhat know...
Exemplary
As an art photographer (not in a professional capacity, but professionally trained), I naturally have an opinion on this. It is, after all, one of the few fields in which I actually somewhat know wtf I'm talking about. Which is not to say that my opinion is authoritative, but this is a field which I've engaged with for the better part of two decades now, and I'm very familiar with its history as well.
My basic opinion is that AI "art photography" can be art, but it's not photography. Even 3D rendering has more in common with photography, because with modern physically based rendering techniques, you can fiddle around with a virtual camera that behaves like a camera and renders light bouncing off objects in an environment in a manner resembling a camera, even if the environment and objects are entirely digital. But that's not really a disqualification of the (potential) value of AI art resembling photography. It simply means that, although we may consider it using some of the techniques and theories relating to photography as art, the process by which it's created is so fundamentally different that we have to consider it a separate art form, just like cinema and photography are related but different mediums.
What really bothers me is the obsession with redefining art to mean "good art", just like "literature" is often synonymous with "great literature". To me, it hollows out the meaning of art to say that only art which rises to a certain level of subjectively assigned artistic merit qualifies. Children make art all the time. Is it good art? Usually not. It has sentimental value to the people who are close to the child, friends and family, but it's not the kind of thing that rises to significance for people who have no relation to the artist. I don't think it's elitist to say that a stick figure drawing of Mommy and Daddy created by a five-year-old is deeply meaningful to the parents, but has no great value outside of that narrow context.
Art isn't just what you like or approve of. There can be such a thing as bad art, or uninteresting art, or derivative art. To declare that art isn't art unless it's both interesting and truly original ignores pretty much the entire history of art, going back to the earliest cave paintings and Venus figurines and passing through all the way to the modern day. Most art isn't groundbreaking. Most art is aesthetically flawed, shallow, or derivative of other art in some way. It can still have value, because it isn't a binary. It isn't either a Da Vinci or van Gogh masterpiece or literal trash; there's a lot of room in between.
Let me give an example of what I personally consider bad, or at least uninteresting and derivative, art. Comedian is an artwork by Maurizio Cattelan. It consists of an ordinary banana duct taped to a wall. It brought a lot of controversy because it sold for millions. As a piece of conceptual art, it isn't about the actual banana or the actual duct tape or even the wall, all of which are replacable. The buyer bought the right to tape any banana to any wall and say it was a work called Comedian by Cattelan. Now, why do I consider this uninteresting and derivative?
In 1917, Marcel Duchamp exhibited his famous work Fountain, which was an ordinary urinal only modified by the addition of a signature, the pseudonym R. Mutt. I'm sure most of you are familiar with it. The work was one of the first so-called readymades, which is the general term for works of art that consist of mostly or entirely unmodified ordinary objects that were not made with the intention of being art, but are presented and exhibited as art in art spaces. Duchamp's Fountain was, at the time, revolutionary. It caused a huge debate about the nature of art, and today, the result of that debate is that works like Comedian are considered art by the art establishment. The art lies not in traditional craft, not in meticulous sculpting, painting, or even photographing. It lies in the concept itself and the provocation and debate it inspires.
Comedian is just a rerun of this concept. It was interesting 100 years ago, because it brought with it new ideas. It has been repeated numerous times since, with all sorts of readymade objects. It's derivative and brings nothing new conceptually, and in terms of aesthetics and craftmanship it's obviously, like all readymades, not exactly impressive (but that was never the point).
Comedian is obviously a satirical work, as hinted in the title. Cattelan isn't a talentless hack. He's made many more traditional pieces of art, albeit with odd and irreverent subjects, such as a sculpture of the pope hit by a meteorite. This is how he was able to sell a banana duct-taped to a wall for millions of dollars. But this particular work is entirely conceptual, and to put it bluntly, the concept is old news. If anything, the social engineering required to sell the work is the real art piece.
Still, even though I think this piece of art is unoriginal and not that interesting, from the perspective of someone who's familiar with the art world, it created a lot of controversy among people outside the art world, which was the point. As such, it's not entirely valueless. It's uninteresting to me as an artist because I've seen it a thousand times before in various guises. It's "not art" and a disgrace to the general public, but that ironically makes it really interesting, a real conversation starter piece.
This is a huge tangent to the topic of AI art, but my point is that the definition of art is very expansive. I can accept Comedian as Arttm (R), but still think it's ultimately derivative and uninspired, or as the kids would say these days, slop.
I don't think Duchamp's work was slop, however. At the time, it genuinely represented something new, which has inspired thousands of artists worldwide. It introduced a view of art as something which was picked out deliberately and presented as art, rather than something that was created using many years of practice in a traditional craft such as painting or sculpting. This has great parallels to photography. Photography is the art of picking out a particular view of the world, during a particular time frame. There is, of course, a lot of photography that's staged, but my favorite kind and one that is very much acknowledged in the art world is the documentary or candid approach, where little to no intervention is performed by the photographer. They choose when, and from where, and how to capture a moment, but the moment itself is not created by the photographer, and in fact, in this kind of art photography, the deliberate creation of a "photographable moment" would diminish it.
AI art is similar in that it's largely an exercise in picking and choosing. The actual act of creation is largely out of the artist's hands. Prompts steer the gen AI, so in that sense it's more akin to staged photography. But if I were to go to an industrial location I'm unfamiliar with and look for some cool industrial machine or structure to photograph, is that fundamentally different from asking an AI to generate images of industrial machines and structures, then picking the best one? In many ways, of course it is. It's based in actual reality, not machine dreams. But in other ways, it's remarkably similar. The artist's hand is involved in the picking out of, and choosing how to present, an object or situation that wasn't created or manipulated by the artist or intended to be art.
In my mind, the interesting thing to consider is how gen AI is used in art, rather than simply that it was used.
I really appreciate traditional craftmanship. I've enjoyed working with old analog cameras even in the age of digital photography. The more cumbersome process places limitations which are often necessary to create interesting art. If you can easily shoot one hundred frames of the same scene, it's easy to be sloppy and end up with none of them being just right. If you can only shoot one, there's a huge risk involved, but it certainly means you'll have to be damn careful and deliberate when you do it. I don't want traditional art to die, and as a documentary-leaning photographer, I really appreciate the deep connection between the medium and reality. It would be a sad day if all we had was AI generated art.
But that doesn't mean there's no place for some of it.
My immediate thought with the subject line was that "AI generated photos" is an oxymoron right out of the gate. Photography is a process, there are several branches it can take, but AI is not...
My immediate thought with the subject line was that "AI generated photos" is an oxymoron right out of the gate. Photography is a process, there are several branches it can take, but AI is not usually one of them, and the photograph is just the final result. "AI generated photorealistic image" is, granted, a mouthful, but more accurate.
Can an "AI generated photorealistic image" be art? IMO yes, absolutely.
Now, regarding that "AI is not usually one of them": I did recently see a fascinating photo of a poorwill flying over the surface of a pond, mouth wide open to skim water off the surface. Album here but you'll have to search for '20700' and scroll; I cannot link directly to the photo. The photographer mentioned that since the intended subjects were bats, the much larger poorwill was cut off, and he used what he termed 'generative expand' to fill in the wingtips. I have to admit feeling a pinch of disappointment there because I feel it would be just as great a photo without the wingtips.
It hadn’t crossed my mind to articulate, but use of gen AI in pursuit of perfecting photography does feel… off. At minimum it has a cheapening effect and it strips the photo of some its value in...
It hadn’t crossed my mind to articulate, but use of gen AI in pursuit of perfecting photography does feel… off. At minimum it has a cheapening effect and it strips the photo of some its value in happenstance.
Of course, that applies to any kind of touching up to a lesser degree, but when believable results can be produced at little cost and no extra required skill, it’s likely to quickly become the norm, crowding out candid and “positively flawed” photos. This feels unfortunate because when flipping back through my album, some of the most interesting photos are those that had been the least planned and edited and most evocative of a small, unfiltered moment in reality.
I'm not a photographer, but it seems to me like the spirit of generative AI in a case like this boils down to: "I didn't get the shot I wanted, but I ALMOST got it, so I'll just nudge it over the...
At minimum it has a cheapening effect and it strips the photo of some its value in happenstance.
I'm not a photographer, but it seems to me like the spirit of generative AI in a case like this boils down to: "I didn't get the shot I wanted, but I ALMOST got it, so I'll just nudge it over the line." It feels a little like "If my lotto number was just one higher, I would've won, so I'll just pretend it was."
Isn't getting the right shot at the right time part of what makes photography special?
I agree that leaving the wings cut off would've been the best outcome for an otherwise nice photo. Knowing that he used generative AI there makes me assume he uses it elsewhere, and also that he is on a journey of increasing usage (because why not?)
This is a phenomenal comment and really describes a perspective I 100% agree with (and it makes my own comments aiming in the same direction look very short and weakly-written by comparison)....
This is a phenomenal comment and really describes a perspective I 100% agree with (and it makes my own comments aiming in the same direction look very short and weakly-written by comparison).
Comedian getting bought by the one dumbass who ate the banana was hilarious enough for me to be glad it exists, though.
When you mention craftsmanship can also be applied to music. We intend to enjoy things more when they are the result of rare talent and hard work, and sometimes even athleticism. A significant...
When you mention craftsmanship can also be applied to music. We intend to enjoy things more when they are the result of rare talent and hard work, and sometimes even athleticism. A significant portion of the interest in music is related to who's the best guitarist or songwriter etc. I just finished watching the movie A Complete Unknown, which is about Bob Dylan. There are many scenes where they show the look on people's faces the first time they first hear something like "The Times They Are A-Changin". From a human perspective, we are very interested in what is possible for us.
Machine generated art can never give us that feeling of awe about what a person is capable of. Similarly, watching Magnus Carlsen play chess is still interesting even though he can't beat a high level computer program.
In line with having different values relating to different art forms, I highly recommend Adam Neely's video essay The Ethics of Fake Guitar. Even just within music, different genres have...
In line with having different values relating to different art forms, I highly recommend Adam Neely's video essay The Ethics of Fake Guitar. Even just within music, different genres have distinctly different values for what makes a song or musician the best representative of their genre or artform, and different ones are more or less vulnerable to being replaced by AI.
I like what you've said about the social engineering being the actual art, and it helped me refine my own idea about this: what is the art of the piece? The banana taped to the wall: If anything,...
I like what you've said about the social engineering being the actual art, and it helped me refine my own idea about this:
what is the art of the piece? The banana taped to the wall: If anything, "the social engineering required to sell the work is the real art piece." So, in a painting, the art can be the technique, or the composition, or the art is the idea. If the creator can explain what the art of the piece is, that'll be a good start.
At one time there were debates over whether photos are art. Surely, some are (art photos are a thing), but the average photo isn’t artistically interesting. I think it’s similar here? There’s lots...
At one time there were debates over whether photos are art. Surely, some are (art photos are a thing), but the average photo isn’t artistically interesting.
I think it’s similar here? There’s lots of meaningless slop out there, but some generated images could attract artistic interest, depending on the process used to make them, how they’re used, or how they’re curated.
Maybe it becomes art through a process of social justification? You need to be able to tell a story about why anyone should care, and that story needs to be accepted by an audience.
Yep. The average ChatGPT-generated image from a simple prompt is like an Instagram photo of a sunset or the Eiffel Tower: easy, cheap, disposable, uninspired. But it's also possible to work with a...
Yep. The average ChatGPT-generated image from a simple prompt is like an Instagram photo of a sunset or the Eiffel Tower: easy, cheap, disposable, uninspired. But it's also possible to work with a variety of tools -- custom models, style transfer, inpainting, layering, curation -- to realize an actual artistic vision, same as with collage, found footage documentary, musique concrète, etc.
The thing that makes it difficult for me to consider AI-generated imagery “art” is the minimal level of user involvement and gacha-like nature of “creation”. The human element comes down to a user...
The thing that makes it difficult for me to consider AI-generated imagery “art” is the minimal level of user involvement and gacha-like nature of “creation”. The human element comes down to a user throwing prompts at the wall until something that pleases them comes out.
Even low-effort photography has a higher level of involvement and intentionality. One has to be there to take the photo, frame the photo to come out the way they want, etc, and the quality of the result is still impacted by the skill of the photographer regardless of the low-effort nature. It’s just wholly a different kind of activity.
Tangentially related, I think many of the earlier forms of generators that produced bizarre, surrealistic images were more artistic overall than modern photorealistic generators. They were still gacha machines but at least brought something new to the table with results that were unexpected and further from the realm of things typically produced by humans (except for those on some kind of substance, maybe).
Yes, the more surreal images were pretty fun until we got used to them. Maybe they'll come back someday? Also, it's a moving target. The "tells" that you can currently use to identify AI art might...
Yes, the more surreal images were pretty fun until we got used to them. Maybe they'll come back someday?
Also, it's a moving target. The "tells" that you can currently use to identify AI art might not be there in a few years, as the generators and tools improve.
Years? With the speed things are improving, I doubt it actually will be that long before we can't spot the difference between human created digital art/photographs vs AI generated ones anymore.
Years? With the speed things are improving, I doubt it actually will be that long before we can't spot the difference between human created digital art/photographs vs AI generated ones anymore.
who, hold on i may be misunderstanding you but my mind is being blown regardless lol—are you saying that the imitation is AI, and that that will always be an imitation of the verisimilitude of...
who, hold on i may be misunderstanding you but my mind is being blown regardless lol—are you saying that the imitation is AI, and that that will always be an imitation of the verisimilitude of actual art? Or are you saying that all art is imitating something & it doesn't matter whether it's AI imitating a painting of a person or a painting imitating a person sitting for a portrait? Because I think maybe it's column A column B.
For a Platonist, all art is an imitation of an imitation. AI art, however, is an imitation of an imitation of an imitation. So, in that logic, human art isn't worth very much, but AI art is worth...
For a Platonist, all art is an imitation of an imitation. AI art, however, is an imitation of an imitation of an imitation.
So, in that logic, human art isn't worth very much, but AI art is worth even less.
Thanks for that—yeah you're not wrong! Maybe "art" is just the wrong word. But full disclosure I also unironically kind of like schticky mashup-type artsy things so Plato would probably have also...
Thanks for that—yeah you're not wrong! Maybe "art" is just the wrong word. But full disclosure I also unironically kind of like schticky mashup-type artsy things so Plato would probably have also been critical of me lol
Not neccesarilly? It absolutely can be and it is significantly easier to create low effort and involvement content than with other methods but it is a tool. The user decides how they use that and...
gacha-like nature of “creation”.
Not neccesarilly? It absolutely can be and it is significantly easier to create low effort and involvement content than with other methods but it is a tool. The user decides how they use that and setups can get very involved.
Wonder how you'd classify something like this, which is where I see the digital art field going once these plugins and tools become more mature and well-integrated into Photoshop et al.
Wonder how you'd classify something like this, which is where I see the digital art field going once these plugins and tools become more mature and well-integrated into Photoshop et al.
The gacha element is still heavy, just more localized, and while intentionality is higher it’s still on the lower end. Overall still a very different creature than photographing or...
The gacha element is still heavy, just more localized, and while intentionality is higher it’s still on the lower end. Overall still a very different creature than photographing or drawing/painting something outright. It’s maybe a bit closer to being art, with the slider moving more in that direction as the artist’s involvement increases, with the highest point on the scale being the artist repainting the entire composition themselves.
AI-based full repaints come with a lot of problems, though. They’re often identifiable by wonky framing, composition, posing, lighting, etc and perhaps worse turn those errors into difficult-to-uproot bad habits in the painter, with any attempt to improve getting canceled out with each repaint.
"Can X be art?". Yes. Art is such a massively wide definition the answer is ALWAYS yes. Which is of course why the article is really just rehashing the same mix of "people just don't like that...
"Can X be art?". Yes. Art is such a massively wide definition the answer is ALWAYS yes.
Which is of course why the article is really just rehashing the same mix of "people just don't like that anyone used AI for this at all" vs "people are using AI to straight up rip off other art styles" (which is still technically art, even if it's also theft).
I'm not really seeing anything substantial to the discussion otherwise.
Yeah it's not really an argument worth having. Mutilating corpses probably counts as art too, but that's not going to be the discussion being had if you start doing it. People find fault with the...
Yeah it's not really an argument worth having. Mutilating corpses probably counts as art too, but that's not going to be the discussion being had if you start doing it.
People find fault with the actions taken to generate the images. Most other merits (or lack thereof), come second to that argument.
Yeah, not only is it pretty hard to argue AI-generated images cannot be art without excluding some things that are very obviously already considered art (e.g., collage), it's just not a productive...
Yeah, not only is it pretty hard to argue AI-generated images cannot be art without excluding some things that are very obviously already considered art (e.g., collage), it's just not a productive argument. It's been over 100 years since Duchamp's Fountain, and if that's art (and pretty much anyone with enough art education to have heard of it agrees it is), obviously so can AI-generated content in the right context.
As with almost all general discussions about "what is art," the perspective tends to come from people without a coherent definition of art to begin with using the label "not art" to arbitrarily exclude things almost solely on the basis of whether the author believes they're good or valuable art.
Yes, and I will repeat myself here: the worthy question is not is this art? but rather is this fair?. I find discussions about what is or is not art tedious. In the case of AI they miss the point...
Yes, and I will repeat myself here: the worthy question is not is this art? but rather is this fair?.
I find discussions about what is or is not art tedious. In the case of AI they miss the point entirely.
"AI"(no such thing currently exists to my knowledge and the bastardization of the term for marketing purposes only creates confusion) is a tool. Are photoshop created images art? It is ultimately...
"AI"(no such thing currently exists to my knowledge and the bastardization of the term for marketing purposes only creates confusion) is a tool. Are photoshop created images art?
It is ultimately created by a person and any qualities that may or may not make it art are imparted by the user.
Only difference here is that the creation process is heavily automated by default and iterative improvements are difficult. Nothing also says additional tools or manual adjustments cannot be used after.
So I made this for someone recently—is it derivative? Yes that's the point. Is it copyright-infringing? Also yes—but did I make any money off of it (& have I uploaded it anywhere at high-res so...
So I made this for someone recently—is it derivative? Yes that's the point. Is it copyright-infringing? Also yes—but did I make any money off of it (& have I uploaded it anywhere at high-res so others can do that)? Nope.
I had the idea, basically collaged a mockup in photoshop with intentions to do a painting, realized the goal of the painting would be for it to look exactly like it already did so why would I do that lol, then had Walmart print it on a canvas for like $25.
So: is it AI? Wellll kinda—how do you decide? I used AI, I didn't make a timelapse or anything so I don't have proof, but I also didn't physically draw or paint anything in meatspace, just on the computer—and I didn't have any original ideas to contribute beyond "cram all the characters onto Noah's Ark together like one of those bizarre anti-smoking videos from 1995." So is it art? idk I like it, I'm prety proud that it turned out exactly how I wanted, but I don't think the "using AI" part makes it "not art;" I think the other factors might make it "not art," but at some point that would be true of all art, which is necessarily derivative & unoriginal—or how about "relevant," "applicable," "empathetic," & all it takes to make that switch is some human element. Yeah someone spamming "Generate" on aiartmachine.com is not making "art," but they're prob not claiming to be doing so, and if they are it's easy to spot.
There are 100% concerns with AI and its rapid & unchecked development, adoption, and sudden pervasiveness, and there are people on here who have told me that it is impossible for anything contaminated AI to be "art," but that seems pretty luddite-esque to me: I think it's one less gate to be kept, in the end, and I think sometimes that's what people are upset about—at least until giant multinational corporations hoover it all up and start charging the masses, then we can argue about whether it should be a right or a public utility to make your friend's dog pictures into Simpsons memes on the internet.
My emotional response is "no, fuck that". But after thinking about it for a few minutes, does the existence of AI art subtract from human art? I don't think so. I would hope that we would be told...
My emotional response is "no, fuck that". But after thinking about it for a few minutes, does the existence of AI art subtract from human art? I don't think so. I would hope that we would be told which art is AI generated, or AI assisted, but we probably won't be soon.
People always want to gatekeep art. When I see a Jackson Pollock painting, I think he seems to have thrown some paint at the wall from a few feet away. But other people enjoy it, and call it art, and that's good enough.
Maybe pretty soon the most valuable art will be something that you watch created in person. Whether it's music or watching someone create a painting. Or maybe value has nothing to do with art anyway.
It certainly at least changes it, doesn't it? In the same way that the existence of photography changes traditional realistic art. One could even say that photography obsoletes it---in which case,...
does the existence of AI art subtract from human art? I don't think so.
It certainly at least changes it, doesn't it? In the same way that the existence of photography changes traditional realistic art. One could even say that photography obsoletes it---in which case, maybe the same could be said of human-made art in the near future?
"obsolete" necessarily implies, if not requires, an imposed uselessness, though; that would only be true if once cameras were invented & affordable & commonplace, no one made art beside...
"obsolete" necessarily implies, if not requires, an imposed uselessness, though; that would only be true if once cameras were invented & affordable & commonplace, no one made art beside photographs. Or if newspapers ceased to exist once radio was invented & adopted, or TV, or cable, or the internet, or social media... is a giant printed newspaper obsolete? I certainly don't see many of them around, but the NYT seems to be doing ok without delivering them to everyone daily.
It seems like the argument is "AI is evil & will destroy our society & take our jobs & stole an idea I had from a dream when I was seven" vs. "AI is the final destination of human achievement," when really it's just AutoComplete from your iphone—super super-advanced autocomplete that will only get better at appearing to be really "thinking," but honestly most of people's concerns about AI seem to be concerns about their own value, intrinsically or economically or both.
And see that's a whole other thing which societally we should definitely be more aware of & concerned about, but wouldn't you know it here's a great thing to both solve the problems & blame the problems on, and an endless supply of people willing to do seminars and provide soundbites for the newspapers internet.
I mean, it is objectively true that, as cameras became increasingly common, the prevalence of photorealistic art diminished. In the modern era, virtually no one make photorealistic art...
"obsolete" necessarily implies, if not requires, an imposed uselessness, though; that would only be true if once cameras were invented & affordable & commonplace, no one made art beside photographs.
I mean, it is objectively true that, as cameras became increasingly common, the prevalence of photorealistic art diminished. In the modern era, virtually no one make photorealistic art anymore---because why bother when cameras exist?
is a giant printed newspaper obsolete?
. . . Kind of, yeah? Virtually no one reads newspapers that way anymore. It's a vestige of a bygone era. If you started society from scratch at our current technological level, newspapers simply wouldn't exist.
I see genAI the same way. Whatever it can do, humans will, for all intents and purposes, stop doing. I think anyone who thinks otherwise is not paying attention to (fairly recent) history.
I don't necessarily agree that hyperrealism art is being done by "virtually no one"—I think that apparent effect is more of a symptom of online prevalence where any individual movement is diluted...
I don't necessarily agree that hyperrealism art is being done by "virtually no one"—I think that apparent effect is more of a symptom of online prevalence where any individual movement is diluted by the enormous mass of just everyone putting all sorts of their own things online all the time; it's like saying "no one likes jazz anymore" when there's probably pound-for-pound more jazz happening every day than ever, but alongside more X as well, where X=literally everything else.
Sorry I wasn't more clear about the NYT—my point was that you can still 100% subscribe to a daily print edition, but the vast majority of people do not because why would they? It costs more, you have a pile of newspaper to do something with every day, no one wants to read all that, etc. It makes sense to move to digital—but A) that doesn't mean everyone will, and B) it hasn't changed the fundamental nature of what the NYT does, just how they present/distribute it. Of course medium will affect content over time, but they are still very much a paper & haven't become Buzzfeed or Barstool or something.
I would add also C) they are still making plenty of money & cultural impact; they have roughly half as many print subscribers in 2025 as they did in 2000 (down to 600,000 including Sunday-only), and their total revenue is down 21% adjusted from what it was then, but they have something like 10x more subscribers. So they've adapted without compromising what they do, which I think is the opposite of obsoletion—in fact since far less of their income is from advertising, I would say they have been able to be more true to what they do than before. Now if you're using the "paper" part of newspapers as the focus then yeah that's not the vibe anymore lol, but it hasn't disappeared & it's a significant enough percentage that it's not comparable to landline phones or something.
To your point re: humans, I think people will stop doing things they don't have to do if they don't have to do them anymore, and people might not pick up a hobby if they're not forced to do it in school for an hour, but also that means that the people who choose to do it will probably really love it, which seems like a net positive to me. Like it doesn't have to be true and bad.
Because you can’t get Jurassic Park or a superhero movie with a similar aesthetic using practical special effects. As computer graphics becomes easier, it gets used in more niches.
why bother when cameras exist?
Because you can’t get Jurassic Park or a superhero movie with a similar aesthetic using practical special effects.
As computer graphics becomes easier, it gets used in more niches.
I may be being pedantic about my favorite thing lol, but the reason Jurassic Park still looks so damn good is the combination of nascent CGI + very very refined practical effects—same as...
I may be being pedantic about my favorite thing lol, but the reason Jurassic Park still looks so damn good is the combination of nascent CGI + very very refined practical effects—same as Harryhausen: whereas Toy Story 4 looks infinitely better than Toy Story 1 or Fantastic Mr Fox v. a Rankin-Bass special, the OG JP, imo, still looks better than any Chris Pratt dinosaur movie ever will.
Yeah, practical effects can often have a nice look to them, too. With enough work, perhaps eventually people will learn to fake that with computers :-) In the meantime, good taste about what to...
Yeah, practical effects can often have a nice look to them, too. With enough work, perhaps eventually people will learn to fake that with computers :-) In the meantime, good taste about what to use when matters.
I don't mean to be disagreeable but I'm not sure about this: I used to lurk reddit a lot, especially fantasy art subs, and it was really common for artists to aim for photorealism. I don't know...
I don't mean to be disagreeable but I'm not sure about this:
In the modern era, virtually no one make photorealistic art anymore---because why bother when cameras exist?
I used to lurk reddit a lot, especially fantasy art subs, and it was really common for artists to aim for photorealism. I don't know why they do it but I always assumed that the creation process was worth more to them than the result. By doing it by hand they're training a ton of techniques and skills. There's also bragging rights and artistic pride.
. . Kind of, yeah? Virtually no one reads newspapers that way anymore. It's a vestige of a bygone era. If you started society from scratch at our current technological level, newspapers simply wouldn't exist.
My parents read the paper every day, and my local gas stations all have full stacks of papers to sell every day. I understand what you mean: it's a trickle of what it once was. But I think it demonstrates that there is still value in older media, and I really don't think AI is going to kill hand made art.
I think art is going to become an increasing but AI driven process for many artists (especially production artists who simply need to meet quotas). I think it's going to put a lot of people out of jobs simply because one person will work much faster by having AI fill in the parts in which they aren't trying to be creative. I expect the human part of the process to move to composition and iterating until it has that je ne sais quoi that will make it effective art.
But I also expect there to always be a non trivial population of non-production artists who simply don't use AI because they don't benefit from it. It can help them make art quickly that approximates their vision, but it can't create their vision.
Maybe my claim as stated is wrong, in the sense that there are, numerically, more people creating photorealistic art than there ever were in, say, the 1600s. I don't even know that I buy that, but...
I used to lurk reddit a lot, especially fantasy art subs, and it was really common for artists to aim for photorealism.
Maybe my claim as stated is wrong, in the sense that there are, numerically, more people creating photorealistic art than there ever were in, say, the 1600s. I don't even know that I buy that, but I'll grant it for the argument's sake.
How about this, then: photorealistic art is, in a general sense, an artform in decline, or even already at the end stage of decline. It was once culturally dominant; but as photography became commonplace, it withered. People may still practice it, but: it does not experience significant innovation; it does not attract top talent; it does not command cultural respect; it has essentially no cultural capital; it does not grant status or prestige or fame; for the vast majority of practitioners, it is not a viable career path; it is not something that most people are interested in or spend time thinking about, except occasionally when an impressive example pops up on their feed. It is a vestige.
I really don't think AI is going to kill hand made art.
See, I'm not 100% sure that this is true, even in a fairly strict sense.
The thing about "photorealistic" art is that, in fact, it does not look exactly like a photograph. At the very least, it admits compositions that are unlikely to ever occur in reality. But it also just looks different. That's part of the joy of looking at it.
But what if photorealistic art could itself be perfectly imitated?
AI art may have a creepy (to me) "AI-ish" look right now, but will it always? Certainly, the AI companies want us to think that, before too long, AI art will be indistinguishable from human-made art. I lack the technical knowledge or insider connections to know whether that's true, but I wouldn't bet against it.
What happens in a world where you can perfectly imitate that style? What happens when so few people are left practicing that techniques are lost, or when so few people are paying attention that all social and financial motivators vanish? What happens when you encroach on the habitat of an endangered species?
People don't just make art in a vacuum. This idea that humans will always find a way to carry on any particular art form is just wrong. Remove the material conditions that allow an art form to flourish, and you can sap its vitality, or even kill it outright. There are plenty of art forms that no longer exist, or exist only in tiny isolated pockets.
It just kinda feels like your missing the point of art... what does "being in decline" mean or what relevance does "innovation" have in an art movement? Art is just the output of a creative...
It just kinda feels like your missing the point of art... what does "being in decline" mean or what relevance does "innovation" have in an art movement? Art is just the output of a creative process orchestrated by a person. Hell, I'm an AI skeptic, but I think art is a place it will continue to flourish. Because its just another tool at an artist's disposal.
Styles of art, like photorealism, are selected by the artist because of taste and process. Photorealism is a bit unique in that historically it has fulfilled multiple use cases: documentarian, portraiture, and artistic. The camera took over documentarian use cases and the majority of portraiture. In that case I can see the argument that photorealism went through a decline around the turn of the 19th->20th century.
As a artistic style photorealism (neo classical) was the defacto format of fine art in the west. This was due to the european academies' desire to arbitrate what was "art" and who could do it. In the 18th-19th centuries en plein air started gaining popularity (I don't remember but I believe part of it was more ready access to pigments for the average person). En plain air birthed impressionism, and abstraction away from neoclassical "photorealistic" art. Expressionism was a further abstraction, followed by surrealism, abstract expressionism, color field, NY abstract expressionism etc. But photorealism didn't disappear. It is the preferred mode of creativity for many artists. There is a distinctly difference experience standing in front of an oil canvas painted by a person and standing in front of a print. So people will continue making it, it isn't in decline anymore than it was in 1890.
This is my take as well. I feel like in modern times we sometimes tie the value of art to it's social value (ie. popularity) or economic value. And by that metric I agree that AI undermines the...
This is my take as well. I feel like in modern times we sometimes tie the value of art to it's social value (ie. popularity) or economic value. And by that metric I agree that AI undermines the value.
I think AI will devalue art for industry, and agree that cameras deprecated photorealistic painting for production purposes. But I don't really care much about that, anyway, and have long felt like the most popular and financially rewarded art is also the least interesting. I enjoy The Avengers, and it's absolutely art, but I consider it's artistic value to be pretty low compared to much less "culturally respected" work. I'm not that bothered by it going to AI beyond the jobs lost (see my final paragraph).
I've always felt like the most interesting art is the stuff that reaches small crowds and is done just because. I don't think AI will interrupt that.
This is now off topic, but I continue to feel like AI presents these problems that are only problems under pressures of capitalism where art needs to be sold because people need to eat. If art were created for the creation, I don't think we'd need to worry about AI. If creating art weren't a job, I wouldn't care if the next Avengers was 90% AI generated.
We seem to be fairly aligned. Though from my perspective art and entertainment has been decoupled. I'm not sure that The Avengers is art. Sure it was created by artists, but how many of them lay...
We seem to be fairly aligned. Though from my perspective art and entertainment has been decoupled. I'm not sure that The Avengers is art. Sure it was created by artists, but how many of them lay claim to their contributions as part of the artistic output versus a contractual exchange of their skills for wages? I think the artist's intent is critical to the evaluation of media as art.
I can consider it art they way I can consider a good heist an art. There's some things in it you can respect in terms of the work required, but it definitely isn't ethical. The example of...
I can consider it art they way I can consider a good heist an art. There's some things in it you can respect in terms of the work required, but it definitely isn't ethical.
The example of re-enacting a historical photo (one that probably happened by had no true evidence) is an odd one to judge, though.
AI is a tool that responds to commands. The authorship belongs to whoever had the aesthetic consciousness to create a new project.
Once we figure out ownership of the data trained on and how thst affects the output, maybe. Many pro-Ai arguments ignore this factor when it comes to generative Ai. I could buy this argument if all art used was CC0.
The answer to "Is X art?" is almost always "yes". I don't feel it's a useful discussion. I think a more useful conversation is one to have with yourself: "Do you enjoy it and/or feel it has...
The answer to "Is X art?" is almost always "yes". I don't feel it's a useful discussion. I think a more useful conversation is one to have with yourself: "Do you enjoy it and/or feel it has artistic value for you". Everyone has their own tastes, and I'm sure many will enjoy AI-generated art.
For me, art that I value has to have a human component involved in its creation, which is more than just random chance. I think there needs to be a vision behind it, an intentionality, so that the finished piece is roughly what was envisaged. Added to that, its value for me increases with human skill. If I enjoy looking at it AND it has been created with skill that I can't personally match, my enjoyment and admiration of it increases more. AI-generated art does not generally fit those criteria... for me.
This is how I like to approach the subject of whether something is art, but I also acknowledge that it's deeply personal and others will have different definitions. And that's ok.
sure, but I am not interested in it at this time. If we consider the prompt to be the art, then sure there is a certain skill for writing prompts, negative prompts, etc that all require an...
sure, but I am not interested in it at this time. If we consider the prompt to be the art, then sure there is a certain skill for writing prompts, negative prompts, etc that all require an understanding of the medium. It isn't unlike writing or anything else in this respect. In the early days of storytelling, the stories were likely, 'we were being chased and there were some acts of bravery... ooga booga' -- but it gets refined over time, and I think the same will happen with AI art.
if we look at something like the holodeck from TNG; this is pretty much AI in an advanced form. That would interest me far more than a dog asking people to like an image that says that Sora should be free for everybody.
Long story short, in its current form I do consider it art but it is very young and not very interesting, but eventually it will be part of a larger thing that has the potential to be very interesting.
Personally, I agree with the existing comments on ethics, but I also think the motivation or end goal of the AI generated material leads to different responses. If a company wants to put up a...
Personally, I agree with the existing comments on ethics, but I also think the motivation or end goal of the AI generated material leads to different responses.
If a company wants to put up a billboard with a creative design, and they use an AI generated image for its entirety, this is a net negative. Not only is source material for many image generators effectively plagiarism, but it is also taking work away from someone who could design something original for them. This also applies for social media posts, brochures, or anything else a business would otherwise hire someone to do.
If an artist wants to use AI to create a representation of their vision, with the end goal of artistic expression, I don't see an issue with this. Art is subjective, and I would look at most AI generated art from a negative lens, but that does not mean it would do the same for others. And potentially, that negative response could be the entire point of that particular art piece.
As an art photographer (not in a professional capacity, but professionally trained), I naturally have an opinion on this. It is, after all, one of the few fields in which I actually somewhat know wtf I'm talking about. Which is not to say that my opinion is authoritative, but this is a field which I've engaged with for the better part of two decades now, and I'm very familiar with its history as well.
My basic opinion is that AI "art photography" can be art, but it's not photography. Even 3D rendering has more in common with photography, because with modern physically based rendering techniques, you can fiddle around with a virtual camera that behaves like a camera and renders light bouncing off objects in an environment in a manner resembling a camera, even if the environment and objects are entirely digital. But that's not really a disqualification of the (potential) value of AI art resembling photography. It simply means that, although we may consider it using some of the techniques and theories relating to photography as art, the process by which it's created is so fundamentally different that we have to consider it a separate art form, just like cinema and photography are related but different mediums.
What really bothers me is the obsession with redefining art to mean "good art", just like "literature" is often synonymous with "great literature". To me, it hollows out the meaning of art to say that only art which rises to a certain level of subjectively assigned artistic merit qualifies. Children make art all the time. Is it good art? Usually not. It has sentimental value to the people who are close to the child, friends and family, but it's not the kind of thing that rises to significance for people who have no relation to the artist. I don't think it's elitist to say that a stick figure drawing of Mommy and Daddy created by a five-year-old is deeply meaningful to the parents, but has no great value outside of that narrow context.
Art isn't just what you like or approve of. There can be such a thing as bad art, or uninteresting art, or derivative art. To declare that art isn't art unless it's both interesting and truly original ignores pretty much the entire history of art, going back to the earliest cave paintings and Venus figurines and passing through all the way to the modern day. Most art isn't groundbreaking. Most art is aesthetically flawed, shallow, or derivative of other art in some way. It can still have value, because it isn't a binary. It isn't either a Da Vinci or van Gogh masterpiece or literal trash; there's a lot of room in between.
Let me give an example of what I personally consider bad, or at least uninteresting and derivative, art. Comedian is an artwork by Maurizio Cattelan. It consists of an ordinary banana duct taped to a wall. It brought a lot of controversy because it sold for millions. As a piece of conceptual art, it isn't about the actual banana or the actual duct tape or even the wall, all of which are replacable. The buyer bought the right to tape any banana to any wall and say it was a work called Comedian by Cattelan. Now, why do I consider this uninteresting and derivative?
In 1917, Marcel Duchamp exhibited his famous work Fountain, which was an ordinary urinal only modified by the addition of a signature, the pseudonym R. Mutt. I'm sure most of you are familiar with it. The work was one of the first so-called readymades, which is the general term for works of art that consist of mostly or entirely unmodified ordinary objects that were not made with the intention of being art, but are presented and exhibited as art in art spaces. Duchamp's Fountain was, at the time, revolutionary. It caused a huge debate about the nature of art, and today, the result of that debate is that works like Comedian are considered art by the art establishment. The art lies not in traditional craft, not in meticulous sculpting, painting, or even photographing. It lies in the concept itself and the provocation and debate it inspires.
Comedian is just a rerun of this concept. It was interesting 100 years ago, because it brought with it new ideas. It has been repeated numerous times since, with all sorts of readymade objects. It's derivative and brings nothing new conceptually, and in terms of aesthetics and craftmanship it's obviously, like all readymades, not exactly impressive (but that was never the point).
Comedian is obviously a satirical work, as hinted in the title. Cattelan isn't a talentless hack. He's made many more traditional pieces of art, albeit with odd and irreverent subjects, such as a sculpture of the pope hit by a meteorite. This is how he was able to sell a banana duct-taped to a wall for millions of dollars. But this particular work is entirely conceptual, and to put it bluntly, the concept is old news. If anything, the social engineering required to sell the work is the real art piece.
Still, even though I think this piece of art is unoriginal and not that interesting, from the perspective of someone who's familiar with the art world, it created a lot of controversy among people outside the art world, which was the point. As such, it's not entirely valueless. It's uninteresting to me as an artist because I've seen it a thousand times before in various guises. It's "not art" and a disgrace to the general public, but that ironically makes it really interesting, a real conversation starter piece.
This is a huge tangent to the topic of AI art, but my point is that the definition of art is very expansive. I can accept Comedian as Arttm (R), but still think it's ultimately derivative and uninspired, or as the kids would say these days, slop.
I don't think Duchamp's work was slop, however. At the time, it genuinely represented something new, which has inspired thousands of artists worldwide. It introduced a view of art as something which was picked out deliberately and presented as art, rather than something that was created using many years of practice in a traditional craft such as painting or sculpting. This has great parallels to photography. Photography is the art of picking out a particular view of the world, during a particular time frame. There is, of course, a lot of photography that's staged, but my favorite kind and one that is very much acknowledged in the art world is the documentary or candid approach, where little to no intervention is performed by the photographer. They choose when, and from where, and how to capture a moment, but the moment itself is not created by the photographer, and in fact, in this kind of art photography, the deliberate creation of a "photographable moment" would diminish it.
AI art is similar in that it's largely an exercise in picking and choosing. The actual act of creation is largely out of the artist's hands. Prompts steer the gen AI, so in that sense it's more akin to staged photography. But if I were to go to an industrial location I'm unfamiliar with and look for some cool industrial machine or structure to photograph, is that fundamentally different from asking an AI to generate images of industrial machines and structures, then picking the best one? In many ways, of course it is. It's based in actual reality, not machine dreams. But in other ways, it's remarkably similar. The artist's hand is involved in the picking out of, and choosing how to present, an object or situation that wasn't created or manipulated by the artist or intended to be art.
In my mind, the interesting thing to consider is how gen AI is used in art, rather than simply that it was used.
I really appreciate traditional craftmanship. I've enjoyed working with old analog cameras even in the age of digital photography. The more cumbersome process places limitations which are often necessary to create interesting art. If you can easily shoot one hundred frames of the same scene, it's easy to be sloppy and end up with none of them being just right. If you can only shoot one, there's a huge risk involved, but it certainly means you'll have to be damn careful and deliberate when you do it. I don't want traditional art to die, and as a documentary-leaning photographer, I really appreciate the deep connection between the medium and reality. It would be a sad day if all we had was AI generated art.
But that doesn't mean there's no place for some of it.
My immediate thought with the subject line was that "AI generated photos" is an oxymoron right out of the gate. Photography is a process, there are several branches it can take, but AI is not usually one of them, and the photograph is just the final result. "AI generated photorealistic image" is, granted, a mouthful, but more accurate.
Can an "AI generated photorealistic image" be art? IMO yes, absolutely.
Now, regarding that "AI is not usually one of them": I did recently see a fascinating photo of a poorwill flying over the surface of a pond, mouth wide open to skim water off the surface. Album here but you'll have to search for '20700' and scroll; I cannot link directly to the photo. The photographer mentioned that since the intended subjects were bats, the much larger poorwill was cut off, and he used what he termed 'generative expand' to fill in the wingtips. I have to admit feeling a pinch of disappointment there because I feel it would be just as great a photo without the wingtips.
It hadn’t crossed my mind to articulate, but use of gen AI in pursuit of perfecting photography does feel… off. At minimum it has a cheapening effect and it strips the photo of some its value in happenstance.
Of course, that applies to any kind of touching up to a lesser degree, but when believable results can be produced at little cost and no extra required skill, it’s likely to quickly become the norm, crowding out candid and “positively flawed” photos. This feels unfortunate because when flipping back through my album, some of the most interesting photos are those that had been the least planned and edited and most evocative of a small, unfiltered moment in reality.
I'm not a photographer, but it seems to me like the spirit of generative AI in a case like this boils down to: "I didn't get the shot I wanted, but I ALMOST got it, so I'll just nudge it over the line." It feels a little like "If my lotto number was just one higher, I would've won, so I'll just pretend it was."
Isn't getting the right shot at the right time part of what makes photography special?
I agree that leaving the wings cut off would've been the best outcome for an otherwise nice photo. Knowing that he used generative AI there makes me assume he uses it elsewhere, and also that he is on a journey of increasing usage (because why not?)
This is a phenomenal comment and really describes a perspective I 100% agree with (and it makes my own comments aiming in the same direction look very short and weakly-written by comparison).
Comedian getting bought by the one dumbass who ate the banana was hilarious enough for me to be glad it exists, though.
When you mention craftsmanship can also be applied to music. We intend to enjoy things more when they are the result of rare talent and hard work, and sometimes even athleticism. A significant portion of the interest in music is related to who's the best guitarist or songwriter etc. I just finished watching the movie A Complete Unknown, which is about Bob Dylan. There are many scenes where they show the look on people's faces the first time they first hear something like "The Times They Are A-Changin". From a human perspective, we are very interested in what is possible for us.
Machine generated art can never give us that feeling of awe about what a person is capable of. Similarly, watching Magnus Carlsen play chess is still interesting even though he can't beat a high level computer program.
In line with having different values relating to different art forms, I highly recommend Adam Neely's video essay The Ethics of Fake Guitar. Even just within music, different genres have distinctly different values for what makes a song or musician the best representative of their genre or artform, and different ones are more or less vulnerable to being replaced by AI.
I like what you've said about the social engineering being the actual art, and it helped me refine my own idea about this:
what is the art of the piece? The banana taped to the wall: If anything, "the social engineering required to sell the work is the real art piece." So, in a painting, the art can be the technique, or the composition, or the art is the idea. If the creator can explain what the art of the piece is, that'll be a good start.
I must ask out of curiosity, not judgement, if you used AI to help with the text in this post.
No. 100% my own words.
Thank you for being a human that is capable of writing so well.
At one time there were debates over whether photos are art. Surely, some are (art photos are a thing), but the average photo isn’t artistically interesting.
I think it’s similar here? There’s lots of meaningless slop out there, but some generated images could attract artistic interest, depending on the process used to make them, how they’re used, or how they’re curated.
Maybe it becomes art through a process of social justification? You need to be able to tell a story about why anyone should care, and that story needs to be accepted by an audience.
Yep. The average ChatGPT-generated image from a simple prompt is like an Instagram photo of a sunset or the Eiffel Tower: easy, cheap, disposable, uninspired. But it's also possible to work with a variety of tools -- custom models, style transfer, inpainting, layering, curation -- to realize an actual artistic vision, same as with collage, found footage documentary, musique concrète, etc.
The thing that makes it difficult for me to consider AI-generated imagery “art” is the minimal level of user involvement and gacha-like nature of “creation”. The human element comes down to a user throwing prompts at the wall until something that pleases them comes out.
Even low-effort photography has a higher level of involvement and intentionality. One has to be there to take the photo, frame the photo to come out the way they want, etc, and the quality of the result is still impacted by the skill of the photographer regardless of the low-effort nature. It’s just wholly a different kind of activity.
Tangentially related, I think many of the earlier forms of generators that produced bizarre, surrealistic images were more artistic overall than modern photorealistic generators. They were still gacha machines but at least brought something new to the table with results that were unexpected and further from the realm of things typically produced by humans (except for those on some kind of substance, maybe).
Yes, the more surreal images were pretty fun until we got used to them. Maybe they'll come back someday?
Also, it's a moving target. The "tells" that you can currently use to identify AI art might not be there in a few years, as the generators and tools improve.
Years? With the speed things are improving, I doubt it actually will be that long before we can't spot the difference between human created digital art/photographs vs AI generated ones anymore.
Perhaps I'm platonist. I don't believe there's a point in which imitation ceases to be imitation.
who, hold on i may be misunderstanding you but my mind is being blown regardless lol—are you saying that the imitation is AI, and that that will always be an imitation of the verisimilitude of actual art? Or are you saying that all art is imitating something & it doesn't matter whether it's AI imitating a painting of a person or a painting imitating a person sitting for a portrait? Because I think maybe it's column A column B.
For a Platonist, all art is an imitation of an imitation. AI art, however, is an imitation of an imitation of an imitation.
So, in that logic, human art isn't worth very much, but AI art is worth even less.
Plato was critical of art.
Thanks for that—yeah you're not wrong! Maybe "art" is just the wrong word. But full disclosure I also unironically kind of like schticky mashup-type artsy things so Plato would probably have also been critical of me lol
For more on that, here's the Wikipedia on Plato's allegory of the cave.
Also this (very long). I haven't read it myself, but SEP is usually great.
I made a conservative estimate because I don't want to contribute too much to the hype. :)
Not neccesarilly? It absolutely can be and it is significantly easier to create low effort and involvement content than with other methods but it is a tool. The user decides how they use that and setups can get very involved.
Wonder how you'd classify something like this, which is where I see the digital art field going once these plugins and tools become more mature and well-integrated into Photoshop et al.
The gacha element is still heavy, just more localized, and while intentionality is higher it’s still on the lower end. Overall still a very different creature than photographing or drawing/painting something outright. It’s maybe a bit closer to being art, with the slider moving more in that direction as the artist’s involvement increases, with the highest point on the scale being the artist repainting the entire composition themselves.
AI-based full repaints come with a lot of problems, though. They’re often identifiable by wonky framing, composition, posing, lighting, etc and perhaps worse turn those errors into difficult-to-uproot bad habits in the painter, with any attempt to improve getting canceled out with each repaint.
"Can X be art?". Yes. Art is such a massively wide definition the answer is ALWAYS yes.
Which is of course why the article is really just rehashing the same mix of "people just don't like that anyone used AI for this at all" vs "people are using AI to straight up rip off other art styles" (which is still technically art, even if it's also theft).
I'm not really seeing anything substantial to the discussion otherwise.
Yeah it's not really an argument worth having. Mutilating corpses probably counts as art too, but that's not going to be the discussion being had if you start doing it.
People find fault with the actions taken to generate the images. Most other merits (or lack thereof), come second to that argument.
Yeah, not only is it pretty hard to argue AI-generated images cannot be art without excluding some things that are very obviously already considered art (e.g., collage), it's just not a productive argument. It's been over 100 years since Duchamp's Fountain, and if that's art (and pretty much anyone with enough art education to have heard of it agrees it is), obviously so can AI-generated content in the right context.
As with almost all general discussions about "what is art," the perspective tends to come from people without a coherent definition of art to begin with using the label "not art" to arbitrarily exclude things almost solely on the basis of whether the author believes they're good or valuable art.
Yes, and I will repeat myself here: the worthy question is not is this art? but rather is this fair?.
I find discussions about what is or is not art tedious. In the case of AI they miss the point entirely.
"AI"(no such thing currently exists to my knowledge and the bastardization of the term for marketing purposes only creates confusion) is a tool. Are photoshop created images art?
It is ultimately created by a person and any qualities that may or may not make it art are imparted by the user.
Only difference here is that the creation process is heavily automated by default and iterative improvements are difficult. Nothing also says additional tools or manual adjustments cannot be used after.
So I made this for someone recently—is it derivative? Yes that's the point. Is it copyright-infringing? Also yes—but did I make any money off of it (& have I uploaded it anywhere at high-res so others can do that)? Nope.
I had the idea, basically collaged a mockup in photoshop with intentions to do a painting, realized the goal of the painting would be for it to look exactly like it already did so why would I do that lol, then had Walmart print it on a canvas for like $25.
So: is it AI? Wellll kinda—how do you decide? I used AI, I didn't make a timelapse or anything so I don't have proof, but I also didn't physically draw or paint anything in meatspace, just on the computer—and I didn't have any original ideas to contribute beyond "cram all the characters onto Noah's Ark together like one of those bizarre anti-smoking videos from 1995." So is it art? idk I like it, I'm prety proud that it turned out exactly how I wanted, but I don't think the "using AI" part makes it "not art;" I think the other factors might make it "not art," but at some point that would be true of all art, which is necessarily derivative & unoriginal—or how about "relevant," "applicable," "empathetic," & all it takes to make that switch is some human element. Yeah someone spamming "Generate" on aiartmachine.com is not making "art," but they're prob not claiming to be doing so, and if they are it's easy to spot.
There are 100% concerns with AI and its rapid & unchecked development, adoption, and sudden pervasiveness, and there are people on here who have told me that it is impossible for anything contaminated AI to be "art," but that seems pretty luddite-esque to me: I think it's one less gate to be kept, in the end, and I think sometimes that's what people are upset about—at least until giant multinational corporations hoover it all up and start charging the masses, then we can argue about whether it should be a right or a public utility to make your friend's dog pictures into Simpsons memes on the internet.
My emotional response is "no, fuck that". But after thinking about it for a few minutes, does the existence of AI art subtract from human art? I don't think so. I would hope that we would be told which art is AI generated, or AI assisted, but we probably won't be soon.
People always want to gatekeep art. When I see a Jackson Pollock painting, I think he seems to have thrown some paint at the wall from a few feet away. But other people enjoy it, and call it art, and that's good enough.
Maybe pretty soon the most valuable art will be something that you watch created in person. Whether it's music or watching someone create a painting. Or maybe value has nothing to do with art anyway.
You're probably thinking of Pollock's drip paintings. He painted them on the floor.
It certainly at least changes it, doesn't it? In the same way that the existence of photography changes traditional realistic art. One could even say that photography obsoletes it---in which case, maybe the same could be said of human-made art in the near future?
"obsolete" necessarily implies, if not requires, an imposed uselessness, though; that would only be true if once cameras were invented & affordable & commonplace, no one made art beside photographs. Or if newspapers ceased to exist once radio was invented & adopted, or TV, or cable, or the internet, or social media... is a giant printed newspaper obsolete? I certainly don't see many of them around, but the NYT seems to be doing ok without delivering them to everyone daily.
It seems like the argument is "AI is evil & will destroy our society & take our jobs & stole an idea I had from a dream when I was seven" vs. "AI is the final destination of human achievement," when really it's just AutoComplete from your iphone—super super-advanced autocomplete that will only get better at appearing to be really "thinking," but honestly most of people's concerns about AI seem to be concerns about their own value, intrinsically or economically or both.
And see that's a whole other thing which societally we should definitely be more aware of & concerned about, but wouldn't you know it here's a great thing to both solve the problems & blame the problems on, and an endless supply of people willing to do seminars and provide soundbites for the
newspapersinternet.I mean, it is objectively true that, as cameras became increasingly common, the prevalence of photorealistic art diminished. In the modern era, virtually no one make photorealistic art anymore---because why bother when cameras exist?
. . . Kind of, yeah? Virtually no one reads newspapers that way anymore. It's a vestige of a bygone era. If you started society from scratch at our current technological level, newspapers simply wouldn't exist.
I see genAI the same way. Whatever it can do, humans will, for all intents and purposes, stop doing. I think anyone who thinks otherwise is not paying attention to (fairly recent) history.
I don't necessarily agree that hyperrealism art is being done by "virtually no one"—I think that apparent effect is more of a symptom of online prevalence where any individual movement is diluted by the enormous mass of just everyone putting all sorts of their own things online all the time; it's like saying "no one likes jazz anymore" when there's probably pound-for-pound more jazz happening every day than ever, but alongside more X as well, where X=literally everything else.
Sorry I wasn't more clear about the NYT—my point was that you can still 100% subscribe to a daily print edition, but the vast majority of people do not because why would they? It costs more, you have a pile of newspaper to do something with every day, no one wants to read all that, etc. It makes sense to move to digital—but A) that doesn't mean everyone will, and B) it hasn't changed the fundamental nature of what the NYT does, just how they present/distribute it. Of course medium will affect content over time, but they are still very much a paper & haven't become Buzzfeed or Barstool or something.
I would add also C) they are still making plenty of money & cultural impact; they have roughly half as many print subscribers in 2025 as they did in 2000 (down to 600,000 including Sunday-only), and their total revenue is down 21% adjusted from what it was then, but they have something like 10x more subscribers. So they've adapted without compromising what they do, which I think is the opposite of obsoletion—in fact since far less of their income is from advertising, I would say they have been able to be more true to what they do than before. Now if you're using the "paper" part of newspapers as the focus then yeah that's not the vibe anymore lol, but it hasn't disappeared & it's a significant enough percentage that it's not comparable to landline phones or something.
To your point re: humans, I think people will stop doing things they don't have to do if they don't have to do them anymore, and people might not pick up a hobby if they're not forced to do it in school for an hour, but also that means that the people who choose to do it will probably really love it, which seems like a net positive to me. Like it doesn't have to be true and bad.
Because you can’t get Jurassic Park or a superhero movie with a similar aesthetic using practical special effects.
As computer graphics becomes easier, it gets used in more niches.
I may be being pedantic about my favorite thing lol, but the reason Jurassic Park still looks so damn good is the combination of nascent CGI + very very refined practical effects—same as Harryhausen: whereas Toy Story 4 looks infinitely better than Toy Story 1 or Fantastic Mr Fox v. a Rankin-Bass special, the OG JP, imo, still looks better than any Chris Pratt dinosaur movie ever will.
Yeah, practical effects can often have a nice look to them, too. With enough work, perhaps eventually people will learn to fake that with computers :-) In the meantime, good taste about what to use when matters.
I don't mean to be disagreeable but I'm not sure about this:
I used to lurk reddit a lot, especially fantasy art subs, and it was really common for artists to aim for photorealism. I don't know why they do it but I always assumed that the creation process was worth more to them than the result. By doing it by hand they're training a ton of techniques and skills. There's also bragging rights and artistic pride.
My parents read the paper every day, and my local gas stations all have full stacks of papers to sell every day. I understand what you mean: it's a trickle of what it once was. But I think it demonstrates that there is still value in older media, and I really don't think AI is going to kill hand made art.
I think art is going to become an increasing but AI driven process for many artists (especially production artists who simply need to meet quotas). I think it's going to put a lot of people out of jobs simply because one person will work much faster by having AI fill in the parts in which they aren't trying to be creative. I expect the human part of the process to move to composition and iterating until it has that je ne sais quoi that will make it effective art.
But I also expect there to always be a non trivial population of non-production artists who simply don't use AI because they don't benefit from it. It can help them make art quickly that approximates their vision, but it can't create their vision.
Maybe my claim as stated is wrong, in the sense that there are, numerically, more people creating photorealistic art than there ever were in, say, the 1600s. I don't even know that I buy that, but I'll grant it for the argument's sake.
How about this, then: photorealistic art is, in a general sense, an artform in decline, or even already at the end stage of decline. It was once culturally dominant; but as photography became commonplace, it withered. People may still practice it, but: it does not experience significant innovation; it does not attract top talent; it does not command cultural respect; it has essentially no cultural capital; it does not grant status or prestige or fame; for the vast majority of practitioners, it is not a viable career path; it is not something that most people are interested in or spend time thinking about, except occasionally when an impressive example pops up on their feed. It is a vestige.
See, I'm not 100% sure that this is true, even in a fairly strict sense.
The thing about "photorealistic" art is that, in fact, it does not look exactly like a photograph. At the very least, it admits compositions that are unlikely to ever occur in reality. But it also just looks different. That's part of the joy of looking at it.
But what if photorealistic art could itself be perfectly imitated?
AI art may have a creepy (to me) "AI-ish" look right now, but will it always? Certainly, the AI companies want us to think that, before too long, AI art will be indistinguishable from human-made art. I lack the technical knowledge or insider connections to know whether that's true, but I wouldn't bet against it.
What happens in a world where you can perfectly imitate that style? What happens when so few people are left practicing that techniques are lost, or when so few people are paying attention that all social and financial motivators vanish? What happens when you encroach on the habitat of an endangered species?
People don't just make art in a vacuum. This idea that humans will always find a way to carry on any particular art form is just wrong. Remove the material conditions that allow an art form to flourish, and you can sap its vitality, or even kill it outright. There are plenty of art forms that no longer exist, or exist only in tiny isolated pockets.
Everything has to end sometime: why not now?
It just kinda feels like your missing the point of art... what does "being in decline" mean or what relevance does "innovation" have in an art movement? Art is just the output of a creative process orchestrated by a person. Hell, I'm an AI skeptic, but I think art is a place it will continue to flourish. Because its just another tool at an artist's disposal.
Styles of art, like photorealism, are selected by the artist because of taste and process. Photorealism is a bit unique in that historically it has fulfilled multiple use cases: documentarian, portraiture, and artistic. The camera took over documentarian use cases and the majority of portraiture. In that case I can see the argument that photorealism went through a decline around the turn of the 19th->20th century.
As a artistic style photorealism (neo classical) was the defacto format of fine art in the west. This was due to the european academies' desire to arbitrate what was "art" and who could do it. In the 18th-19th centuries en plein air started gaining popularity (I don't remember but I believe part of it was more ready access to pigments for the average person). En plain air birthed impressionism, and abstraction away from neoclassical "photorealistic" art. Expressionism was a further abstraction, followed by surrealism, abstract expressionism, color field, NY abstract expressionism etc. But photorealism didn't disappear. It is the preferred mode of creativity for many artists. There is a distinctly difference experience standing in front of an oil canvas painted by a person and standing in front of a print. So people will continue making it, it isn't in decline anymore than it was in 1890.
This is my take as well. I feel like in modern times we sometimes tie the value of art to it's social value (ie. popularity) or economic value. And by that metric I agree that AI undermines the value.
I think AI will devalue art for industry, and agree that cameras deprecated photorealistic painting for production purposes. But I don't really care much about that, anyway, and have long felt like the most popular and financially rewarded art is also the least interesting. I enjoy The Avengers, and it's absolutely art, but I consider it's artistic value to be pretty low compared to much less "culturally respected" work. I'm not that bothered by it going to AI beyond the jobs lost (see my final paragraph).
I've always felt like the most interesting art is the stuff that reaches small crowds and is done just because. I don't think AI will interrupt that.
This is now off topic, but I continue to feel like AI presents these problems that are only problems under pressures of capitalism where art needs to be sold because people need to eat. If art were created for the creation, I don't think we'd need to worry about AI. If creating art weren't a job, I wouldn't care if the next Avengers was 90% AI generated.
We seem to be fairly aligned. Though from my perspective art and entertainment has been decoupled. I'm not sure that The Avengers is art. Sure it was created by artists, but how many of them lay claim to their contributions as part of the artistic output versus a contractual exchange of their skills for wages? I think the artist's intent is critical to the evaluation of media as art.
I can consider it art they way I can consider a good heist an art. There's some things in it you can respect in terms of the work required, but it definitely isn't ethical.
The example of re-enacting a historical photo (one that probably happened by had no true evidence) is an odd one to judge, though.
Once we figure out ownership of the data trained on and how thst affects the output, maybe. Many pro-Ai arguments ignore this factor when it comes to generative Ai. I could buy this argument if all art used was CC0.
The answer to "Is X art?" is almost always "yes". I don't feel it's a useful discussion. I think a more useful conversation is one to have with yourself: "Do you enjoy it and/or feel it has artistic value for you". Everyone has their own tastes, and I'm sure many will enjoy AI-generated art.
For me, art that I value has to have a human component involved in its creation, which is more than just random chance. I think there needs to be a vision behind it, an intentionality, so that the finished piece is roughly what was envisaged. Added to that, its value for me increases with human skill. If I enjoy looking at it AND it has been created with skill that I can't personally match, my enjoyment and admiration of it increases more. AI-generated art does not generally fit those criteria... for me.
This is how I like to approach the subject of whether something is art, but I also acknowledge that it's deeply personal and others will have different definitions. And that's ok.
sure, but I am not interested in it at this time. If we consider the prompt to be the art, then sure there is a certain skill for writing prompts, negative prompts, etc that all require an understanding of the medium. It isn't unlike writing or anything else in this respect. In the early days of storytelling, the stories were likely, 'we were being chased and there were some acts of bravery... ooga booga' -- but it gets refined over time, and I think the same will happen with AI art.
if we look at something like the holodeck from TNG; this is pretty much AI in an advanced form. That would interest me far more than a dog asking people to like an image that says that Sora should be free for everybody.
Long story short, in its current form I do consider it art but it is very young and not very interesting, but eventually it will be part of a larger thing that has the potential to be very interesting.
Personally, I agree with the existing comments on ethics, but I also think the motivation or end goal of the AI generated material leads to different responses.
If a company wants to put up a billboard with a creative design, and they use an AI generated image for its entirety, this is a net negative. Not only is source material for many image generators effectively plagiarism, but it is also taking work away from someone who could design something original for them. This also applies for social media posts, brochures, or anything else a business would otherwise hire someone to do.
If an artist wants to use AI to create a representation of their vision, with the end goal of artistic expression, I don't see an issue with this. Art is subjective, and I would look at most AI generated art from a negative lens, but that does not mean it would do the same for others. And potentially, that negative response could be the entire point of that particular art piece.