
What If People Like AI-Generated Art Better? (christies.com) 156
Christie's auction house notes that an AI-generated "portrait" of an 18th-century French gentleman recently sold for $432,500. (One member of the Paris-based collective behind the work says "we found that portraits provided the best way to illustrate our point, which is that algorithms are able to emulate creativity.")
But the blog post from Christie's goes on to acknowledge that AI researchers "are still addressing the fundamental question of whether the images produced by their networks can be called art at all." . One way to do that, surely, is to conduct a kind of visual Turing test, to show the output of the algorithms to human evaluators, flesh-and-blood discriminators, and ask if they can tell the difference.
"Yes, we have done that," says Ahmed Elgammal [director of the Art and Artificial Intelligence Lab at Rutgers University in New Jersey]. "We mixed human-generated art and art from machines, and posed questions — direct ones, such as 'Do you think this painting was produced by a machine or a human artist?' and also indirect ones such as, 'How inspiring do you find this work?'. We measured the difference in responses towards the human art and the machine art, and found that there is very little difference. Actually, some people are more inspired by the art that is done by machine."
Can such a poll constitute proof that an algorithm is capable of producing indisputable works of art? Perhaps it can — if you define a work of art as an image produced by an intelligence with an aesthetic intent. But if you define art more broadly as an attempt to say something about the wider world, to express one's own sensibilities and anxieties and feelings, then AI art must fall short, because no machine mind can have that urge — and perhaps never will.
This also begs the question: who gets credit for the resulting work. The AI, or the creator of its algorithm...
Or can the resulting work be considered a "conceptual art" collaboration — taking place between a human and an algorithm?
But the blog post from Christie's goes on to acknowledge that AI researchers "are still addressing the fundamental question of whether the images produced by their networks can be called art at all." . One way to do that, surely, is to conduct a kind of visual Turing test, to show the output of the algorithms to human evaluators, flesh-and-blood discriminators, and ask if they can tell the difference.
"Yes, we have done that," says Ahmed Elgammal [director of the Art and Artificial Intelligence Lab at Rutgers University in New Jersey]. "We mixed human-generated art and art from machines, and posed questions — direct ones, such as 'Do you think this painting was produced by a machine or a human artist?' and also indirect ones such as, 'How inspiring do you find this work?'. We measured the difference in responses towards the human art and the machine art, and found that there is very little difference. Actually, some people are more inspired by the art that is done by machine."
Can such a poll constitute proof that an algorithm is capable of producing indisputable works of art? Perhaps it can — if you define a work of art as an image produced by an intelligence with an aesthetic intent. But if you define art more broadly as an attempt to say something about the wider world, to express one's own sensibilities and anxieties and feelings, then AI art must fall short, because no machine mind can have that urge — and perhaps never will.
This also begs the question: who gets credit for the resulting work. The AI, or the creator of its algorithm...
Or can the resulting work be considered a "conceptual art" collaboration — taking place between a human and an algorithm?
How you define art (Score:2)
That pretty much sums it up. AI collects what already exists, but doesn't go into new territory. People mostly like what's familiar, and the new, most creative Art often makes them uncomfortable at first.
AI, as it exists now, is a long way from what I would c
Re: (Score:2)
"doesn't go into new territory" It seems to when it hallucinates although the hallucination generally seems to be derivative. That is, putting stuff together in surprising (to us) ways. Then so do politicians so one cannot ascribe any active intelligence behind it. It is sort of like that special CEO who comes up new catch phrases to explain whatever whizzy thing s/he is doing, you know it isn't well thought out but it is hard to quantify it if it isn't contradictory.
Re: (Score:2)
Hallucinations are not creating new things. It is creating things that are false or nonsense.
Re: (Score:2)
And art is not? If I draw Jesus riding a Utahraptor while shooting laser guns at aliens that would arguably be art (save for the part where I can't actually draw), but asking an AI to come up with a similar picture is not? I agree that AI-generated works shouldn't be art, but I'm also aware that it's an emotional response on my part.
Re: (Score:2)
The missing bit seems to be intentionality. The way the latest AI image generators work there is no possibility for creativity. The model doesn't deliberate and make decisions about how to respond to the prompt. The prompt just affects way in which the model performs its singular task: removing noise. That this results in recognizable images related to the prompt is absolutely astonishing, but it is an entirely mechanical process, not a creative act. Your prompt might be creative, but the output is deci
Re: (Score:2)
I haven't messed with AI directly myself. I enjoy doing my own writing and I have no interest in making video or photos.
If two people give the exact same prompt to the same AI model, they should get the exact same result. Is this correct?
It could be argued that my unique set of prompts used with a specific model produces said creative work. If you don't know my unique set of prompts, then you wouldn't get the same produced work as I would.
If that's true, then I would say a person using AI to create a photo
Re: (Score:3)
If two people give the exact same prompt to the same AI model, they should get the exact same result. Is this correct?
No, because the AI model works on random noise, and the random noise pattern is different each time. Either chosen based on a random number generator, or the person using the tool can specify the noise pattern. You can even paint certain noise patterns, such as a rough outline of shapes, and cause the imagen to fill in draw the more detailed lines and shade the noise pattern you've g
Re: How you define art (Score:2)
Re: (Score:3)
The missing bit seems to be intentionality.
It doesn't really matter. You are way off base, because intentionality OR deliberation is absolutely not required for creativity.
Sometimes humans deliberate. Other times human creatives simply play around and try MANY random different things until they arrive at a result that they like. Many works are created by accident with zero intentionality.
Of course elements of intentionality come up in the way a prompt is chosen to be written, And which if any generatio
Re: (Score:3)
Last century's debates on Dadaism should inform us here. In "Fountain" the urinal continues to do nothing other than exist as object made for another non-art purpose, but few would argue that "Fountain" isn't art today.
I don't see what differentiates an industrial product an artist attaches a label to from a some AI software (an industrial product) an artist (prompt author) utilizes. If you are concerned that model/software isn't the same as the output we could look Warhol's Factory work the output images
Re:How you define art (Score:5, Insightful)
Re: (Score:3)
It's fundamentally incapable of coming up with a truly original idea,
No more so than humans are. The only truly original ideas actually come from the natural world itself. Based on your criteria Humans are fundamentally incapable of coming up with a truly original idea. Artists only extrapolate on what they have observed and learned from nature.
Re: (Score:2)
No more so than humans are. The only truly original ideas actually come from the natural world itself.
I dispute that, but also this sounds suspiciously like a no true Scotsman in the making.
Take maths for example. We've come up with all sorts of weird shit with no real counterpart in nature.
Anything involving infinity of any sort, including irrational and transcendental numbers. Now some of them, such as pi are inspired by true events as they say in Hollywood, but what about the Louisville and Champernowne
Re: (Score:2)
Re: AI Banksy, severely limited (Score:2)
Re: (Score:2)
All art is always derivative. (That's a very slight exaggeration. Extremely slight.)
The problem with AI art is that the AI doesn't have any inherent sense of style, it's entirely determined by what it was taught. (OTOH, different systems *can* emphasize different aspects of what they have been taught. Which might be called a rudimentary inherent sense of style.)
That being the case, the quality of the AI art is entirely determined by the process of selection. But this is largely true of all art anyway.
Art is not what people like (Score:2, Insightful)
Art, from the Greek poiesis, means what is produced. The artist makes art, not the people. And given that the people in general do not like to think but rather prefer to feel, I really do not care what they like. They've given us Taylor Swift and Donald Trump, both 'artists' who thrive on giving the people what they want.
Yes, I am an elitist. Poor, but an elitist none the less.
And for the love of God, it does not beg the question. It raises it. The former is an informal fallacy [wikipedia.org]; the latter a figure of speec
Re: (Score:2)
Have you read the Jeff Hawkins book?
Re: (Score:2)
Jeff Hawkins is a crackpot.
Re: (Score:2)
NAK
Re: (Score:2)
I disagree.
The AI collects and AVERAGES what already exists. So not only is new material not made, but it reinforces material that is easy to access (eg public domain, BSD-licensed or similar), and thus becomes familiar. This is why you see the Wizard of Oz, and Alice in Wonderland in so many stories, because these are public domain. Hence everyone knows these stories and are familiar.
So over time what is going to happen is AI's will absorb more AI slop made by other AI's, and this will result in a form of
Re: (Score:3)
You see the same stuff over and over again because there really are only 7 basic plots anyway.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Sadly, Hollywood is on par with AI for just regurgitating almost the exact same shit over and over again. They don't even bother to spin up new characters in new places doing the same basic plots. Nope, just remake the movie ever so slightly.
The older you get, the more obvious this is. You've seen it before and it's not new.
It's also why you are better off advertising to kids. They'
Two reasons why this argument doesn't work (Score:3)
First of all, art isn't defined as saying something or expressing an artists feelings. If it was, then much of what we see in museums wouldn't be considered art. I'd also say that some of what tries to say something is more pretension than actual art.
But the more important things is that currently AI doesn't create by itself. It's directed by people, and judged be people. The resulting image is one that went through a lot of iterations of getting somewhat that pleases the human creator. In this respect if w
Re: How you define art (Score:2)
The ability to share ideas across time is something that defines humans. Best we can tell, the rats have some pretty amazing lives, but no Louvre to share the rat experience with future generations.
Re: (Score:2)
It's music if someone, anyone, listens to the sound pattern and appreciates it. It's music to me if I'm the one who appreciates it. Wolves don't make music, they communicate with sounds. It only becomes music when a human appreciates it.
To claim that the wolves communication via sound is music when no human hears it is to claim knowledge of the interior mental states of the wolves.
Re: the Art of Wolves (Score:2)
Other cultures who live amongst wolves and have deep connections with dogs such as the Inuit have a different take on this.
The people the AI creators ripped off that is who (Score:5, Informative)
The algorithm doesn't really deserve credit for creating anything, nor does the creator of the algorithm. Is it is the creators of all the works of art the creators of the AI model ripped off that really deserve the credit. Most in a case like this are probably long since dead, but using an AI to create a digital painting like this is one step removed from plagiarism. In addition, anyone who pays nearly half a million dollars for something AI models will likely be able to generate in a few minutes a few years down the road is probably some combination of naive, uneducated, or insane.
Re:The people the AI creators ripped off that is w (Score:5, Insightful)
By that measure, an artist who paints a work doesn't deserve credit for it. Only the artists that came before him in the model he was trained with. In meatspace, their art history textbooks for example. So as you say, anything a trained human artist creates is also one step from plagiarism by your definition
Until people can delineate a clear distinction between AI trained on publicly visible art and humans trained that -same- way on the very -same- art, we are going in circles.
“Good artists copy, great artists steal” - Pablo Picasso
Re: (Score:3, Insightful)
AI and humans learn in very different ways. A human does not digest a huge dataset, produce random output, assess if that output is desired, tweak the parameters a little and repeat until the result is "good enough". Also, a human learns from a significantly smaller dataset than generative AIs.
Most humans also pay hommage to those that they were inspired by. When is the last time you saw an AI tell you who they were trained on? Oh, you don't because the AI companies don't want to acknowledge that the traini
Re:The people the AI creators ripped off that is w (Score:4, Insightful)
A human does not digest a huge dataset, produce random output, assess if that output is desired, tweak the parameters a little and repeat until the result is "good enough".
They kind of do. They absorb an enormous dataset (everything they have ever seen, with emphasis on the things they have looked at the most) and the output is often not what is intended and has to be repeated and refined to attain a desired result.
What humans don't have is a deterministic model. That makes what we are doing possibly both more and less impressive in different ways, and also arguably less violating of copyright (which is whatever the legal profession decides it is, if you include lawmakers.) If you can get the machine to spew out exactly the same results which are arguably infringing (not going to get into that here) just by using the same numbers every time, doesn't that make it more like "a machine" than we are? And doesn't that make what they're doing fundamentally different? Which, of course, we already know it is. Even though what it's doing has some similarity to what we do, it is definitely not using the same process even if it were only because the hardware is so very different.
Re: (Score:2)
For what it's worth, I don't believe humans are deterministic, but I don't like your argument that requires art to be the result of a non-deterministic process any better. Some human had to prompt the AI to
Re: (Score:3)
Umm, they hell they don't. Go listen to any genre of music from a specific era and you'll here bands that all sound really similar. People hear something on the radio, like it, and then start playing their instruments to sound like what is popular. Had they never heard that prior sound, they wouldn't be trying to adapt it to themselves.
It's not to say no artist out there is going against the grain, but a huge amount of our "culture" is us copying other popular people with minor deviations. Vast majority of
Re: (Score:3)
Honestly, this is exactly what my 2 year old does when speaking to adults. His "good enough" result is "adults can understand what I'm saying".
I cannot imagine how anyone that has even met a toddler wouldn't immediately identify that this is _precisely_ what they do.
Re: (Score:2)
Until people can delineate a clear distinction between AI trained on publicly visible art and humans trained that -same- way on the very -same- art, we are going in circles.
So - you're under the impression that a person can read some art history text books and then just start spewing out duplicates and remixes of those images directly? You think that's how people actually work?
Here's a distinction: human beings aren't computer programmed algorithms. If we were, photocopiers and cameras wouldn't have been invented in the first place, because we'd all have the ability to glance at anything from a friend's birthday cake to a dollar bill and then recreate it with pixel perfect cla
Re: (Score:2)
Until people can delineate a clear distinction between AI trained on publicly visible art and humans trained that -same- way on the very -same- art, we are going in circles.
The brain is a learning machine, not a computer. Computers are deterministic, brains are not.
I really wish AI proponents would stop saying that AI works exactly the same way our human brains do. That's idiotic.
Re: The people the AI creators ripped off that is (Score:3)
Business Art sells and if it doesn't, it goes out of business.
Something had to replace crypto mania. Hello AI!
Buy now! Amaze and impress your friends!
art is (Score:3)
If you like it, it's art.
Re: (Score:2, Insightful)
That's the crafts part of arts and crafts. You may appreciate the aesthetic, but not everything is art, nor does it need to be.
Re: art is (Score:2)
art and communication (Score:2)
I never said things could not be beautiful. A sunset or waterfall is beautiful, but it's not art.
Not the creative person decides if it is art, but the recipients.
You and Leo Tolstoy agree. But I can't really use your definition (or Tolstoy's) and still have the word "art" retain any meaning.
Ultimately we use agreed on definitions to quickly communicate ideas. If your definition is off the rails, you should probably be explicit in what you mean and leave that word alone going forward.
Re: (Score:2)
Not quite. You don't need to like it. But if you are emotionally moved by it, then it is art. (I'm not sure that's quite a complete definition. Perhaps it should also include rationally moved. Consider "This is not a pipe".
giving credit to whom? (Score:3, Interesting)
Wow, won't you even consider giving credit to the artists who the AI is plagiarizing?
> > This also begs the question: who gets credit for the resulting work. The AI, or the creator of its algorithm...
Re: (Score:2)
The artists being plagiarized, the person guiding the AI to do the plagiarizing, definitely not the AI itself (it is just a tool) and not the people who created the AI.
A bu ch of guys used hammers to build my house. Neither the hammers nor the people who designed the hammers have any claim on having built my house.
Re: (Score:2)
The artists being plagiarized
Can you really call it plagiarism when the influence of any individual work on a generated image is likely less than a single pixel's worth of information?
the person guiding the AI to do the plagiarizing,
As I argue elsewhere, while a prompt might be a creative work, the resulting image is decidedly not. If you won't accept that, then consider the contribution of our hypothetical prompt writer in terms of information. It doesn't make for much of a claim.
not the AI itself [...] and not the people who created the AI.
Obviously.
Rather than ask who should get the credit, we should be asking why anyone should get the cre
Re: giving credit to whom? (Score:2)
Re: (Score:2)
I will sell you Picasso's A Home Depot Hammer for a mere $500k. It's a steal! :-)
Re: (Score:2)
who gets credit for the resulting work
No one.
The individual artists whose art was used as training data individually have so little influence on any generated image that to call it plagiarism is beyond absurd.
I don't see that the creators of the model have any claim. The whole point of AI is that humans don't need to design the resulting algorithm. Of course, I don't see how they would have any claim to the output even if it was a traditional algorithm. If I use a CAD tool to layout a board from a schematic, would the creators of that algori
Re: (Score:2)
People are just mad because AI generating "good enough" imagery is eating their lunch. If I'm a company that needs new business cards, or new menus, or a new billboard advertisement, I can just sit around and prompt away until I get "good enough".
Before, I would have had to hire someone to do the same process. Probably took longer and cost more and the results are just as subjective as the AI's output.
So of course people are scared and upset.
On the other hand, a painter can produce paintings and if other pe
Not Surprising (Score:4, Insightful)
However, coming up with something truly new and unique that has never been done before and that turns out to be a sensation is probably beyond AI at the moment...and currently also beyond Hollywood too.
Re: (Score:2)
I wouldn't say it's beyond Hollywood (well, maybe) or AI. AI could reassemble things in a way no one else would ever consider doing. We could call that "new".
Hollywood is afraid to take any risk.
There is very likely all sorts of "original" works of art being created, but they need to be discovered and there are at the very least, 10s of millions of artist producing stuff. Hard to get noticed and go viral, so to speak.
Re: (Score:2)
I wouldn't say it's beyond Hollywood
I would at the moment - the current cost of big, blockbuster movies has become so inflated that they can't afford to take risks on someting entirely new and innivative because the downside if they get it wrong is too large. Ironically, AI may actually help here: by potentially making it much cheaper to make films it allows them to once again take risks.
Re: (Score:2)
Can you name anything that is "truly new and unique"?
It doesn't matter what anyone says, you'll claim any similarity make something derivative. Take Jackson Pollock, for example. As far as I'm aware, his work was completely without precedent, "truly new and unique". But he did make paintings and we've had paintings for ages, so you'll dishonestly reject that example.
Humans come up with truly unique things all the time. It's how progress is made. Babbage's analytical engine wasn't the first mechanical aid to calculation, but it was the first general purpo
Re: (Score:2)
I mean, depends how you measure time :)
Re: (Score:2)
It doesn't matter what anyone says, you'll claim any similarity make something derivative. Take Jackson Pollock, for example. As far as I'm aware, his work was completely without precedent, "truly new and unique". But he did make paintings and we've had paintings for ages, so you'll dishonestly reject that example.
Humans come up with truly unique things all the time. It's how progress is made. Babbage's analytical engine wasn't the first mechanical aid to calculation, but it was the first general purpose calculating machine ever designed. That concept represented a radical shift in our understanding so great that we remember him to this day even though he was unable to realize his vision. It was, by any reasonable definition, "truly new and unique", but other machines existed, so it doesn't count, right?
You can hand-wave away anything I can propose or point out some obscure earlier take, but ultimately every innovation has to have some origin. The world didn't spring into existence last Thursday.
I have no idea what "truly new and unique" is supposed to mean which is why I asked for examples. By your descriptions doesn't seem you know either.
Is this purely a subjective term where you'll know it when you see it or is there objective criteria that can be applied to discriminate between "truly new and unique" and "recombines its training data"?
What if anything prevents a brute force search or simple genetic algorithm from creating something "truly new and unique"?
What prevents a diffusion model from c
Re: repetition (Score:2)
Re: (Score:2)
Can you name anything that is "truly new and unique"?
Yes: for books I'd say Lord of the Rings and Frankenstein are good examples. While these novels may have taken some ideas that existed before them they added completely new ideas that went far beyond anything else at the time to create something that, at least for me, meets the bar of something completely new and unique.
Current LLMs are incapable of that kind of leap although you can get such leaps from AI when they can be judges based on purely objective criteria - such as AI's that have found and expl
All the training on other peoples work has to stop (Score:2, Insightful)
The major objection to AI is that it is trained on other artists work. Then perhaps we need to ban all work by anyone or anything that is influenced in any way by prior copyrighted work. /s for you m
We also train humans the same way, and that would simply have to stop.
If an artist cannot certify that they have never taken an art class, looked through an art book, been to a gallery, or ever talked with another artist, their work should be banned in the same way as AI trained the exact same way.
Its only fair.
Re: (Score:3)
And so are other artists. It's called art school. We charge money for it and the original artists don't see a penny unless they happen to be the teacher.
Sounds like we are just being racists against computers. We judging them for doing things in a different way then we do them and that's bad, mmkay.
Every single living person has taken in their lifetime's experiences. Every tree seen, every car, every person, all the colors, the conversations and so on. All these things are inputs, including the movies, musi
Re: All the training on other peoples work has to (Score:3)
But if quality can be generated enmass? (Score:3)
Re: (Score:2)
>"That might lower the prices of all art works along Christie's and the art markets bottom line."
Meh, who cares?
People will pay what they want for what they want. There are human-created works of "art" that are just junk to me that fetch ridiculous prices, I don't see this as that much different. As long as there is no fraud involved (like claiming some AI-generated thing *was* created by a human, especially some well-known artist), then it shouldn't matter.
That said, generally, most people pay high pr
Re: (Score:3)
I imagine high end art is used as a money laundering vehicle.
This will get me modded down, but Hunter Biden's art was totally worth $225,000, right? https://nypost.com/2022/12/10/... [nypost.com]
Re: But if quality can be generated enmass? (Score:2)
Does mass production lower the value of a Rothko or Lucian Freud? Hardly. The Art market is as much about provenance as it is aesthetics and there's a huge industry of experts who argue the veracity of art.
I think that one of the more
I like the ease we can create it. (Score:3)
Thus far the major use case I have found for AI generated art is to share some randomly generated image with my wife. That is, to create something ephemeral which will only be seen by one other person, just for the amusement value of the image. Like generating an image of a Ring-billed Gull wearing a top hat.
But increasingly I'm seeing AI art used elsewhere--and to be honest I'm not a fan. But for some reason or another we've decided that all articles must be accompanied by a 'hero image' (and to be honest, I'm not a fan; I prefer words over images or videos), so it makes sense we'd see a proliferation of low-quality 'bullshit' art filling those 'hero image' layouts because that's what the design required, not because the design was created for usability.
To be fair, I prefer human-generated art, only because part of the value of art is that it's a form of expression: a statement the artist was making, and the best art is art where the artist was trying to say something in a particularly clever way.
Re: (Score:2)
Wait ... are you saying you don't want to see the same generic, slightly cartoonish, 'artwork' plastered at the top of page? You'd rather have an actual artist produce something clever and interesting that provides some useful insight or otherwise prepares the reader for the article (or whatever) that follows?
I'm shocked. Shocked, I say.
Re: (Score:2)
Re: (Score:2)
Depends! Don't you find the Escheresque collection of mismatched fingers and the garbled pseudo text to be helpful. But no wonder she can't get a job, typing with 7 fingers where half of them sort of emerge as the gap between the others must be awful. No wonder her text comes up as random scribbles.
capitalist propaganda (Score:4, Interesting)
Problem is why those portraits were created. (Score:2, Insightful)
The majority of 18th century 'gentlemen' portraits were not created to say say something about the world.
They were created for various reasons, including things like:
1) Social Standing
2) Dating (yeah, they 'photoshopped' the crap out of these)
3) Gifts for loved ones
4) Historic legacy
5) In payment for support. - give $$ to a school, get your portrait hung there.
An AI could do many of these function as well if not better than a human.
Re: (Score:2)
The majority of 18th century 'gentlemen' portraits were not created to say say something about the world.
A good portrait can say a great deal, regardless of the reason they were commissioned.
[...] An AI could do many of these function as well if not better than a human.
You seem to be confused.
AI art (Score:4, Interesting)
As we are constantly exposed to AI art, I think we will just feel more alienated by it. It won't resonate with us because it won't be human.
Re: (Score:2)
But are that things set in stone or just not considered yet? I bet the majority of people like music on the grid like they prefer symmetric faces, still some find it more than uncanny. But who says that an algorithm needs to have the drum sound perfect each time? Nothing is easier than introducing a bit of noise. Of course, the complicated part comes in, when you want to add noise to the data to make it interesting but avoid disturbing it too much at the same time. But that's exactly a task that an AI syste
Re: (Score:2)
Re: (Score:2)
Animal emotions are much easier to understand and emulate than rational thought. It's literally insect-tier level of difficulty vs something only humans can do.
Re: (Score:2)
Instinct is not much more than reacting to rules that are hard to describe in words. AIs are very good at learning unwritten rules just from given examples.
For the creativity you now just need to limit following too much rules. Yes, the AI can also learn to deviate from the perfect drum, but you won't want it to deviate in a grid pattern. So again there needs to be some source of randomness.
Fiction and poetry are mostly dead ... (Score:2)
... as in written by humans that is. Just try it. They are shuffling stuff really well so they aren't in fact plagiarizing anyone and for these use cases it doesn't matter (well, actually it's a requirement) if they hallucinate.
Re: (Score:2)
Good fiction uses story to tell us something about the human condition. If all you care about is the plot and setting and don't mind the same crap endlessly recycled, I guess AI fiction is the thing for you. Just don't be surprised when you don't seem to get the same value out of reading that you did before.
Stolen art (Score:2)
Main thing (Score:2)
The main thing this ignores is that in a healthy society, art is used not just for visual stimulation but also for communication between human beings. AI art may have the former, but it will never have the latter. If AI visual art becomes more appealing, it only shows the sick nature of modern society.
ridiculous (Score:2)
The AI has no agency. That's like giving a chevy credit for skidmarks that were made while doing donuts. You could say the truck maker had some impact on the skids and arcs but that we don't typically do so, their AI applied to the sale and distribution of the tires and truck but the truck itself is a tool of the operator and the skidmarks the operator's creative expression if anyone's.
People don't like AI "art" better (Score:2)
...some people follow trends and like whatever is new. Some believe that they can sell it to the next trendloving sucker for a higher price
Using AI to generate "art" is a misuse of the tech
We already have artists who make art really well
We need AI to do things we can't do
Art is just Money Laundering (Score:3)
"Art" (Score:2)
I define "art" as something that can be created only by living, thinking, feeling beings.
What is being hailed as "AI art" now is not actually creations, but "mash-ups" of existing artworks.
So what's new? (Score:2)
'“Good artists copy, great artists steal” - Pablo Picasso
Re: (Score:3)
I define "art" as something that can be created only by living, thinking, feeling beings.
Any term one can have debates about what it means and personal definitions is useless in conversation and no friend of the man who seems to simply communicate effectively and not waste his time.
You vastly overestimate the taste of the average p (Score:2)
Re: (Score:2)
>"You vastly overestimate the taste of the average person"
The average person doesn't have almost half a $millon to throw away on some AI-generated piece of "art" that is neither special nor rare. Like others have said, it sounds kinda fishy.
What if there's a new fad tomorrow? (Score:2)
AI art is new and amazing, however you feel about it, it is still frequently impressive. It is far more frequently crap, but that doesn't change the feeling of awe when you type in a phrase describing some piece of artwork that didn't exist and it poops out more or less (or sometimes almost exactly) what you were imagining. This accessibility where even unskilled people can get online and type some sentences, and have someone else's computers produce these images for them makes it a fundamentally new experi
They don't (Score:2)
The scheme is really simple. You buy a piece of "art" for a hefty sum. Your purchase of that art establishes a market for it and the value naturally rises. Because art is subjective you can declare the value to be whatever you want and as long as the IRS plays along get away with it. You then donate
Re: (Score:2)
>"When rich people buy art it's a tax dodge and a scam. For the most part the loophole has been closed in the United States but with the recent administration change I suspect it's going to get opened up again."
You correctly described exactly what Hunter Biden was doing- creating "art" and then he could rake in lots of money from people buying his silence/obedience/political favors/etc. But then you try to say this is the kind of stuff the newly-incoming administration would favor/do while ignoring the
Re: (Score:2)
gentleman = moron (Score:2)
"gentleman" obviously has more money than brain. quite literally.
there will be a market (Score:2)
I'll bet you a beer that the wifey-poo bought it at the same time as the couch and carpet, because the colour scheme matched.
That is not a terrible reason to buy it, but
You (I) can't expect people to have goo
"AI" Cannot Make Art (Score:2)
They won't (Score:2)
Christie's auction house notes that an AI-generated "portrait" of an 18th-century French gentleman recently sold for $432,500. (One member of the Paris-based collective behind the work says "we found that portraits provided the best way to illustrate our point, which is that algorithms are able to emulate creativity.")
It's called novelty, just like NFTs it's fun for the ultra-rich who can waste some cash to enjoy it while it's cool, but it will wear off quickly.
One way to do that, surely, is to conduct a k
People like Coldplay and voted for the Nazis (Score:2)
No meaning (Score:2)
Great art captivates people because it communicates underlying meanings or emotions. We muse about what da Vinci was feeling or thinking when he pained Mona Lisa, or where the next Banksy art will show up.
AI art doesn't have any underlying emotion or meaning. It's the equivalent of art students painting copies of other people's masterpieces. Some art students do it really well, but the art isn't theirs, it's somebody else's. The students' rendering doesn't add anything to it or make people think.
People migh
Re: (Score:3)
These people have way too much money to spend if this is the kind of crap they are buying. Look at how bad they look.
Eh, you can say exactly the same about a lot of human art - especially modern "art".
Here's [bbc.co.uk] an article about a urinal ("created" by French artist Marcel Duchamp in 1917 and put up for auction in 2002); the article bemoans the fact that it sold for "only" $1,185,000. Yeah, the object has some historical context, but the context appears to be "look how dumb art buyers are". Or, for a more modern example, a banana duct taped to a wall sold recently for over 6 million [npr.org].
If I had 6 million to throw away I'd take an
Re: (Score:2)
"Only fine art is art." Do you have any idea how stupid you sound?
Re: (Score:3)
You don't know what "woke" means, do you? I'd feel bad for you, but you're an asshole.