The ethics of AI art

Post Reply New Topic

From what I’ve seen of AI art so far it appears to be almost exclusively images of warrior-women with unfeasibly large breasts. While there is obviously an infinite appetite for such trash, it’s ultimately not very challenging on any level.

This stuff is not produced by horny robots with a brush in their hand, it’s some pathetic guy accessing a programme, setting some parameters, and considering himself an “artist”.

I see no shame in doing anything for money.

I think what people hate the most is being confronted with the actual worth of their creative effort relative to what they think that should be. The French Impressionists started out making quick, cheap paintings for middle class people to hang on their walls. Renoir started out painting porcelain, fans, and window shades. Renoir eventually came to earn his living doing portraits, but single paintings didn’t earn him that much. He just painted a lot and people were willing to pay for it. He made so many paintings that by the time he became famous, art students would bring him fakes to “touch up” and sign his name to. Now Renoir paintings fetch a high price at auctions, but Renoir himself will never see a dime of it.

It’s like back when I was in college I knew a few people working on their MBA degrees, and then they couldn’t figure out why they didn’t get hired by Fortune 500 companies and make six figures. Art and money is like that. Music and money is like that. Anything worth having is worth working and earning. Art and music that earn a lot for the artist are rare and in demand. Otherwise, you make your actual living in the cheap consumable business where we mere mortals live.

The only REAL shame isn’t making money. The REAL shame is abandoning things you love because it doesn’t make a lot of money. Always be true to yourself, always do what you love. Excellence is always rewarded when you play the long game. In the meantime, do what it takes to earn a living.

I share your sorrow about people doing art they don't themselves like, and I suspect you've hit on the best touchstone there is for determining whether or not an artist is selling out. Moneymaking, I see that as rather more complicated than "is making money shameful?" and expect I could fill a book with my nuanced thoughts on the relationship between moneymaking and morality. As for excellence, how would anybody determine that about art?

I think “selling out” is just something unsuccessful people say. It’s sour grapes. What makes me happy? Music. I prefer creating new music. I prefer performing in front of audiences. And if that means writing arrangements of other peoples’ music or playing for church, I’m grateful for that. And when I see people who make a lot of money doing big stuff, I can either be bitter and jealous or I can celebrate their achievement. Being bitter is unproductive, though, while cheering on others is inspiring and motivating. So complaining about sellouts who sell out arenas when that’s what I wish I did is really just a waste of a life. I can achieve what I want in life in so many different ways I’ll never feel like a lesser person if I don’t “make it.”

I view excellence in objective terms. Excellence is simply establishing an objective standard and meeting or exceeding that standard. Ideally, that mark should be set high. Art and music are propagandized in subjective terms so much it’s confusing to try to say what good or bad art even is. So I look at it in objective moral terms.

Briefly: What is “good” is that which holds objective human values in the highest esteem—freedom, reason, and life, but really all that upholds the best in human living. For us, life is more than mere existence. We don’t dig burrows like some animals do. We build houses, we decorate our living space to reflect ourselves, and we insist on having electricity and running water. Those things are nice but not necessary for survival. Human living is more than that or it’s not considered living. I realize that this seems like reducing existence to the collection of stuff, but that would be a gross misunderstanding. The arguments pro and con would be best in a different thread.

So to apply moral good to visual art, where is humanity in the picture? Renoir’s portraits place humanity and the joy of life and light in the center. The beauty and proportions of the female form AS IT IS is glorified, no fat shaming whatsoever. His women are healthy, glowing. In paintings of crowds, like Moulin de la Galette, people are happy and carefree.

Much of impressionist art is landscape art and still life. Humanity is absent. These paintings may offer a window into a beautiful world in a room that needs a little brightening up, but the elevation of humanity is at best neutral, at worst the natural world is best with us missing from it.

Romantic art and music is more consistent in its realism and elevation of human values.

Modernism and Postmodernism are more opposed to human values. Exceptional artists such as Frank Lloyd Wright did much more to advance art and architecture beyond Romanticism. But the standard was lowered in the post-war brutalism of Bauhaus. Buildings are “machines for living,” with cheap and easily accessible materials like concrete making them objects you wouldn’t really miss if bombs fell on them. Some people might even say if these building were destroyed you would be doing humanity a favor.

If you blew up a FLW, it would be a national outrage. Nobody cries when an urban Section 8 housing project gets demolished. FLW represents forward effort in human living. What does Bauhaus-inspired high-rises represent? Poverty, organized crime, the decline of the nuclear family, greedy absentee landlords, CPS visits, corrupt police, racism, failing school systems, intrusive government… I’m just skimming the surface here, but FLW is the achievement of the American dream. The alternative is the acceptance of despair and being trapped in a regressive culture.

One is life, the other is death. One is good, one is evil. Most of us live somewhere in between, but we desire a life and lifestyle that leans in the direction of FLW. Raphael, Da Vinci, Renoir, Bouguereau represent moral good and excellence. Minoru Yamasaki, Edvard Munch, Theodor Adorno might represent the absolute worst in regressive thinking and great evil.

To be fair, at least with regard to Yamasaki, modernism came with the best of intentions. Yamasaki wasn’t a terrible designer. Pruitt-Igoe had been altered from the original plan to cut costs, and the threat of forced integration scattered all residents except those who felt unable to move. So maybe Pruitt-Igoe wasn’t designed to fail. It was BUILT to fail.

Excellence is about what is best on individual terms and exceeding the standard. Great art is worth saving and grieving when it’s gone. Great art is timeless. It elevates humanity and improves our lives. It creates heroes. It achieves. What did Pruitt-Igoe accomplish? Ghetto. Produced victims. Generated poverty. Ended in dynamite.

So as to defining excellence in art, I think that’s a good place to start—what is RIGHT and GOOD in our world? How do we take human life and living and make it BETTER for the individual? How do we set high standards and blow past them? How do we celebrate those who achieve greatness? I don’t expect everyone to share my taste in art and music. I think that would be tragic. But I do think that some basis in reality, something that exists outside the mind, is the only place to start.

If subjective good is all that matters, why did Pruitt-Igoe fail? Actually…IMO I think a deeper question is this: Why was Kurt Cobain such a SUCCESS? Nirvana’s music was terrifically tragic and self-destructive. The popularity of Nirvana speaks negatively to the attitudes of an 80’s/90’s youth culture that GLORIFIED self-destruction. In defining excellence, it’s also important to assess how societies can reject it and prefer degradation. Stalinism, National Socialism, the Branch Davidian cult, Heaven’s Gate, Aum Shinrikyo, Al Qaeda. Self-destructive artistic movements seldom ideate ACTUAL destruction, but they do for ideas what evil cults and social movements do human existence. I would say that while artistic excellence feeds human rationality, destructive art poisons it. Toxic art and culture is vastly more evil and dangerous than National Socialism since the seeds of all kinds of evil germinate within them. Don’t forget Hitler was a visual artist, and National Socialism was a highly visual movement. And to the NAZI, the things they were doing were for the greater good. A subjective world view cannot define excellence nor morality in any meaningful way. Without relating moral and artistic excellence to reality, it doesn’t exist. There is no basis for comparison. And I believe that lowering the standard or abandoning all standards, we are a danger to ourselves. Art often reflects are standards and values, and communism, national socialism, and the kinds of thinking that spawned trash like Pruitt-Igoe is indicative of an abandonment of human excellence and achievement.

I'm sure it can be sour grapes in some cases, but not all. When I was in a band, we got a chance to sell out, compromise our values, or whatever it's rightly called. We were doing a certain style of music that we loved doing. Then somebody familiar with the working men's club circuit in the UK advised us that we should try that. We looked into it, and it soon became clear that we were perfectly capable of going that way, but that we'd have had to change our act an awful lot, and that we didn't like anything that the club-goers liked. There wouldn't have been anything all that technically hard about it, in fact in some ways it would have been easier because we'd have been playing down to the audience who wanted something a lot simpler than what we were doing, but the choice of songs and the entire presentation stood for everything we personally disliked about the performing arts. So we decided to forego the potential for money and prestige, and to just keep going our own way whether we ever made money out of it or not, because having fun and sticking to our preferred ways was more important to us than being paid for it.

I've got the highest regard for bands who ended up making money from the same genres we were into. The amount they're paid doesn't really enter into it, and I've usually no idea whether they got super-rich or just broke even. Naturally I feel a certain envy that they were able to do it full-time while I always had a "normal" job to distract me from my greater interests, but it's a friendly kind of envy. My disdain is reserved for the behaviour of those who cynically play stuff they dislike in the hope of getting popular or rich, whether or not they succeed with it, except in cases where it's the only means at their disposal to get what they need to do the stuff they genuinely enjoy doing.

I quite like Munch's art. As well as Da Vinci's and Renoir's.

Certainly if an artist aims for a particular result it's conceptually possible to objectively measure the success or failure of the final result of the work, like throwing a dart at a board when you've declared which number you're trying to hit, and in that sense I agree with you. As for objective morality, the only thing there that makes much sense to me is the principle of harmlessness, i.e. that if somebody is doing no harm, it's not logically valid to condemn their behaviour, that victimless "crimes" aren't immoral acts. I might be highly suspicious of anybody "making" huge amounts of money by art or by any other means, on the grounds that I suspect they can't get rich without making other people poor, which is harm, but that's a whole other argument that depends on whether or not the zero-net-gain thing is correct, and probably beyond the scope of this thread.

All I can say about what is good art and what is bad art is that I know what I like. I've said that I feel a certain disdain for the behaviour of "artists" who voluntarily do the kind of art they dislike simply in order to make money, but that's purely my own disdain, and if somebody else doesn't mind, then it's not for me to judge their attitude, just that it won't change my attitude to it, and it doesn't matter how many people tell me a piece of art is "great," it won't make it float my particular boat. I suppose that makes me a type of aesthetic relativist.

This is ultimately what it boils down to. Though there is art I don't like that I can still appreciate by virtue of the work that went into it. I'm not a big fan of Cubism, but Picasso did put a lot of work into the Cubist paintings and did some interesting things. But I'd much rather have a Rothko or a Breughel the Elder in my room.

Personally, I like to think of design as the practical side of art. It is about rules, regulations and most importantly function over form. They say that good design is invisible. At least in commercial design. The average person isn't going to notice when a design is good, but they'll notice when it doesn't work. I'm a graphic designer. A lot of what I do is visual problem solving. I often work with very strict branding rules which don't allow for creativity.

Usually a client will come to me and say 'Hey! This is my brand, here's a twenty to forty page list of what you can and can't do. You see, I have a problem, my brand usually works great, but in this specific context it looks terrible. What can we tweak to make it work?' or 'Hey, we're looking to create a partnership with another brand, but our brands look quite different, which is good in a way - we want to remain separate commercial entities, yet we want to visually represent the fact that we're working together in this specific circumstance and we want it to not look awful - help?'

Although I do get the odd client who doesn't really know what they want and they just have a vibe. Sometimes an MS Paint file. I don't mind that so much, it can be a somewhat entertaining challenge. Most brands don't have such in depth branding guides, but I tend to work with fairly well established brands that are very specific and will absolutely go on a rant if you don't meet the exact guidelines. I'm under a non-disclosure agreement, so when it comes to my portfolio, I use my own concepts to represent the type of work I can do for others. I don't want to face the wrath of certain companies.

Frankly I'm somewhat surprised that AI art didn't come for graphic design first. I'm under no pretences that a robot couldn't do my job. Although admittedly it would struggle with understanding the more vague comments you get from clients - e.g. make it more dynamic, I want it to pop, and so on. You could in theory teach that to a robot, you could make it read theory books and teach it what certain phrases tend to mean. Although I think clients typically prefer talking to a human, since sometimes they like to ask for feedback from the designer.

Granted, not all graphic design is the same. I tend to class graphic design under two categories - freehand (allows for creativity, you can make your own designs) or strict.

Then there's creative expression art. AI largely seems to be targeting digital background design, especially steampunk backgrounds, along with illustrations of humans. As previously pointed out in this thread, it's mainly going for anime styles with disproportionate women. Sometimes men.

Two major problems are currently holding AI back. AI, at least the generators that are currently being used, don't understand concepts such as race. It sees a human and goes - Oh! I know this one! Then proceeds to throw together the images it has in its database. Which has led to people who aren't white being portrayed as such by the AI. It also doesn't recognise physically disabled people as people (such as amputees). Since it doesn't have enough reference images being fed to it that have disabled people, it gets confused and makes a different guess.

Another issue is that it doesn't understand body parts. You get weirdly twisted fingers, or ears on people's backs or extra fingers in weird places. If you're looking at a piece of art that's almost picture perfect, but you look at the hands and it has the most messed up hands you've ever seen - it's probably been made by an AI.

I make creative expression artwork for fun. This is fairly different from the work I do to earn money.

With me I'm so stubbornly autonomic that I can barely appreciate anything I don't personally like. When I'd only seen abstract offerings by Picasso I was convinced for years that he couldn't make a "proper" painting if he tried, but later I saw some of his earlier stuff and had to revise my opinion and admit that he was indeed competent.

Interesting.
It makes sense to me why this would be unsettling. Imagine having spent years training to be the finest chef. Years of sweating in kitchens and creating adored and celebrated dishes, with feedback about how wonderful your cooking is, appearances of masterchef, as well as making that a large part of your identity. Uniqueness, a special ability, something that only the best trained can do, and suddenly some f**k invents a replicator.

So, this is to some degree about value, and identity, perhaps? Suddenly, the cost of creating quality, within a domain dropped. This is indistinguishable from reduction in value (if supply increases, price must drop to match for demand, something like that). The way I see this, ai effectively reduces the value of artists work, as seen by everyone else, including the artists themselves. I say this not to antagonise any artists, but to try to get a feel for what psychological processes are driving this, and why there would be such distress.

And then, the deeper rumblings of doubts about human value, soon follow, perhaps. What is next? Art is supposed to be something only human minds can create, because creativity was supposed to be something impossible to program. The same cheapening, and reduction in value could be seen to then apply to creativity itself, and shortly thereafter the human mind.

It's a bit of an illusion though. Ai came about through an unimaginably long and tedious process, genius upon genius iteratively improving technology all the way back from the first impressions in the sand, across multiple domains. Although when the ai's can be created by just pushing a few buttons, it can seem like it's "easy". A sort of hedonic adaptation to tech, if you will.

Doesn't alter the loss of something precious to the artists though.

^
I can relate to a lot of that. Technology has a nasty habit of de-skilling artisans and rendering their long efforts a waste of time. It happened to me in a small way as the creator of music recordings because of computers. I'd slaved away for years to get impressive results from my own skills as a musician and from affordable, real musical instruments and recording equipment which took a lot of hard work and ingenuity to do the job right. Then suddenly anyone could achieve better just by pressing a few keys on a computer. Luckily for me the technology had serious artistic limitations - it turned out that perfect timing and tuning etc. didn't sound as artistically pleasing as the manually-created thing, and computer programs struggled to add authentic humanisation to the music. But these days the difference isn't so large and may have vanished for those who can afford the best equipment. I suppose it's always been that way though. The artist who can afford better paint and brushes will get a better result without being a better artist.

I’m on the opposite end of the spectrum. I prefer embracing technology.

I would like to point out some VERY important limitations of tech, though, to make a larger point. Sample libraries have expanded and evolved to allow computer musicians to create frighteningly realistic symphonic and cinematic works. The limitation, though, is that samples are “set in stone,” so the composer is strictly limited to the sounds and techniques that have been sampled. I can always smell a young composer who is too attached to the past and wants to be the next Beethoven because the mentality is “if I just write traditionally and use professional sounds, my music will be great.” The music consistently falls flat. What are they doing wrong? They’re writing for real instruments but counting on tech to know what they want without giving the computer enough instructions to mold and shape the music and get realistic results. Also, a sampled clarinet and a real clarinet are NOT the same instrument. AT ALL. Composing for a sampled violin and a real violin are as different as composing for a real violin and a real piano. NOT THE SAME. But inexperienced composers will pretend that they are. Experienced composers using samples will write with the samples in mind, will add expressions in real-time, and expect (reasonably) that live musicians can imitate the mock-up for best results. It’s much easier for human performers to imitate machines than it is for machines to imitate humans.

I’ve played guitar just enough to understand basically how guitar works in modern recordings. I’m a terrible PLAYER, but I understand the theory. So I can record samples guitar tracks that are almost indistinguishable from the real thing. I’m also really good with sampled drums. But with making realistic drum tracks, I know what MIDI performances do to drum sounds. It’s…horrible, to say the least. But then it’s a learning experience when you can create compelling drum tracks without machine-gunning the sample kit. Think less is more, and make very conservative use of sampled rolls and flams. It might not be as flashy as the real thing, but at least you can trick the listener into thinking it’s real. At worst, it doesn’t draw attention to itself.

The exceptions to the rules of making convincing compositions with computers is electronic music. Here, machine-gun snare rolls are expected and even desirable. Auto-tuned vocals are the norm. Real pianos and guitars sound out-of-place. Just because music “sounds digital” or robotic doesn’t mean it’s less expressive.

It is true that computer music is much easier in a lot of ways than doing things the old school way. Instant gratification. What’s wrong with computers isn’t that it’s easy. What’s wrong is that the ease of computers gives the illusion that users don’t need any theoretical background or years of hands-on practice to get big-studio results. Things like EQ and compression work the same way as their analog counterparts. But until you have a lot of practice with mixing and mastering, unless you have a good library of reference songs and practice listening as much as recording, you’ll never have the quality results as the pros.

AI art can't be that beneficial to clients other than offering a huge variety of random generated images

Followers of Rupert Sheldrake will know of his theory of "morphic resonance". If applied to art then human generated art is much more complex than AI in that it taps into something that was first postulated by Carl Jung called the collective unconsciousness.

Morphic resonance has been shown through experimental evidence that people are capable of "tuning in" to other conscious experiences in what could be translated to "telepathic connections". Therefore human generated art not only provides a more complex product but by "tuning in" one can (according to morphic resonance theory ) generate a brand, logo that is more connected (or makes more of a connection) to other people's conciousness.

The idea of the collective conciousness has been around for at least 100 years. Consciousness is not something you can proves exists but there are anecdotal signs people have telepathic abilities. The concept provides some ammunition (even if not rock sold) to artists and designers why human generated art may be more desirable to a client than something created by an AI bot.

Oh yes, my brief anecdote wasn't the whole story of my experiences with technology. I got a music computer when such things were relatively rare, simply because it was cheaper than buying a multi-track recording machine that could give me more than 4 tracks, and initially was only interested in its MIDI capabilities in as far as it could replace my very limited drum machines with something more versatile. But once I'd started to explore the extra features I was soon using them.

Interesting topic....I should also mention the cutting edge work of Michael Levin
https://en.wikipedia.org/wiki/Michael_Levin_(biologist)

He is not only on the way to a Nobel prize but through his research on the intersection of AI, nueroscience and cognition he will likely demonstrate that collective minds (literally minds everywhere) are a reality
https://www.frontiersin.org/articles/10 ... 68201/full

Morphic resonance (Rupert Sheldrak and Carl Jung) on steroids where every living being has a mind that is connected. It lends itself to George Lucas's concept of "the force" and the conciousness that Nikola Tesla famously claimed to be able to tap into.

Again I don't pretend to be an expert but given his published work is only months old and weirdly he is also working with artificial intelligence. Stay tuned folks this might have ramifications for AI art.

Or to put it simply - collective intelligence makes human generated art more connected to the needs of a client than AI could ever do.