I think this topic drifted more into a debate over AI art in general than I intended. I don't disagree that it's a very problematic tool creatively, economically and ethically. But I do see some uses for it that I think respect those issues while not writing the technology off completely.

For those who have spoken out against it, are there any applications for AI art that you would consider acceptable or ethical, or is it a hard "no" in all situations?

An experienced digital artist knows how to compose a scene, what colors work together, how to make something look good, how to draw hands.

This line (in terms of visual media) is determined by whether or not someone needs to know what makes something visually appealing in order to make something visually appealing.

...

AI image prompting removes the need for theory, for all you need is to give it instructions on what it is you want, the only bit of theory around it is how to tell the AI what you want. The AI is what's providing most of the visual appeal. This is the line where tools turn to automation.

I agree that composing a prompt doesn't require any theory or knowledge of what makes something visually appealing. However, it doesn't preclude it either. Here is an example prompt given by an AI art image generator:

High detail RAW color photo professional [...] close photograph of a ((blonde woman standing)), in a city or a field, [...] matte, pastel colors, backlighting, depth of field, natural lighting, hard focus, film grain, (3d), ray traced, rendered, photographed with a Sony a7 III Mirrorless Camera, by photographer

The specificity of the prompt shows that whoever wrote it knows both theory and the traditional tools of the craft very well, and they're going to have a lot more control over the end result because of it. Yeah, you can also generate an image from a simple prompt with no knowledge, but you're not going to have much control over the result. You might like what comes out, but like I said previously, is that any different from an amateur photographer taking a snapshot of a landscape they think looks good without any knowledge of lighting, framing or composition? You can recognize that something looks good visually without understanding why it does.

Similarly, Jackson Pollock can throw paint at a canvas and we consider it high art because he had a great understanding of traditional forms of art before he started experimenting with splatters. If you look at Pablo Picasso's early works, you can see he both knew how and had the skill to create realistic portraits before he began experimenting with cubism. If I haphazardly chucked paint at a canvas or started sketching subjects at weird angles, I could probably create something very similar but nobody is going to respect it because A) I'm just aping someone else's style, and B) I don't have the knowledge and experience to justify such experiments.

E. E. Cummings famously didn't use "proper" capitalization or punctuation in his poetry, but he knew all of the grammatical rules and how to compose traditional poetry, so we can appreciate him subverting those rules and forms. Whereas if a child did the same thing, we'd correct them because they're doing it out of ignorance rather than experimentation. And we're definitely not going to fault a typewriter or word processor for allowing someone without knowledge or experience of writing to make such rudimentary mistakes even if the end result resembles something created by an expert.

All of this is to say that it's not the tool or the medium you should be concerned about. It's the person using them that matters.

I still do not get why so much of the conversation is focused on whether it requires skill or not.

...

Some people's valuation of art is increased by knowing the skill and effort required to produce it, but it's not a requirement for all people's appreciation. I love hand-drawn and stop-motion animation a lot more than CGI, but that doesn't mean I dislike CGI animation entirely. It can be done well, and I actually love when it's used to replicate hand-drawn/stop-motion animation. Because, even if it's easier to produce, it shows that the animators understand and appreciate those traditional methods enough to replicate them. The movie Klaus is a great example.

And my appreciation of AI art largely comes from the skill of humans required to create machine learning systems rather than the AI itself.

I think other parts of it are much more dangerous: like ChatGPT being ascribed some sort of communicative intent or "knowledge", when in reality, it is literally just a text generator without any knowledge of the world. The prospect of using ChatGPT and other similar "large language models" for anything but the party trick it is right now is incredibly dangerous. It just puts words behind each other. And of course I don't feel threatened as a programmer or writer by ChatGPT because it might have been trained on my code or writing.

This is a big issue as well. I don't think we're close to creating artificial intelligence that is sentient or sapient. But I also don't think the first time we create such AI will be intentional or immediately recognizable. It's going to emerge from the machine learning systems we're working on today for other applications, and it's going to go unnoticed for a long time before we're aware of it. So I can kind of sympathize with that Google engineer who was fired because he thought their chat bot had become self aware, but I do think he's very naive and easily duped. Natural speech engines will get very convincing long before they reach self awareness. His reaction came from a place of caring, though, and I hope that's how we'll respond when truly self-aware AI emerges.

However, we also need to be wary of emergent intelligence that doesn't resemble our own. That's going to be the most difficult to identify.

There's a conversation in It's Always Sunny in Philadelphia that fits here.

Frank: "So that painting I bought from you is worthless?"

Art Dealer: "Of course not! It meant something to you. It's worth exactly what you paid."

Frank: "I wanna sell it back."

Art Dealer: "Okay, in that exchange it would only be worth what I would pay for it, which is nothing."

---

But seriously, I've seen art valued in a few different ways. There is commissioned art, where the artist will either charge an hourly amount or a flat fee for the finished work. I personally think a flat fee for any commissioned work is incredibly dangerous to the artist, because the client may refuse the finished product and demand changes which adds to the time spent without the artist being able to charge extra for it. If you're billing hourly, the client should know the quality of your finished work from your portfolio and the artist should have to give an estimate of how many hours it will take.

But then there is art that is created by the artist of their own volition and then sold as a finished piece. It's high risk to the artist because maybe no one will pay what they're asking for it. If no one buys it at the asking price, the artist will be forced to either lower the price or continue to wait for someone who may or may not show up to buy it (and in the meantime not make any money from it). But if someone sees a piece of art, sees the price tag, and decides that the price is acceptable, then the art was worth at least that much to the buyer.

The amount of time spent on the piece doesn't really matter in this case, unless that's something the buyer factors into their own decision to purchase it. And no one else's valuation of the piece matters. They can criticize and argue all they want (like in that instagram thread you linked) but they're wasting their time because the only two people that it should matter to are the artist and the buyer.

The one thing I asked my English teacher each year in high school was if we'd get to do any creative writing that semester. The answer was always "no."

In elementary school, the English/Language Arts classes always included a creative writing portion and I loved it, but it was totally absent in high school. I did eventually get to take a creative writing course at university, but the professor was a bitter old nutjob with tenure who did absolutely zero teaching and spent most of his time griping about fanfiction and his one-sided rivalry with another professor that everyone loved.

I most recently watched the new Pinocchio movie that came out last year. No, not that one. Not that one either. Definitely not that one! The one by Guillermo del Toro! It was really good. I love the stop-motion animation and the slightly creepy visual style and story. It feels like a classic Tim Burton movie.

I knew someone would give me this reaction.

Oh my goodness, that is so funny.

Geocities definitely felt more advanced than Expage. If I recall correctly, you couldn't just edit the whole site with Expage. There were separate fields for the header, body, and footer that you could edit separately, and you were stuck to choose some of their limited options for graphics, dividers, songs, etc. It was still great fun though. I met my best friend on that website, we've been friends for over 20 years, because I stole the color of his background and that made him upset.

The front page of Expage looked like this:

This brings back so many memories. Some of the other limitations I remember were you couldn't use HTML, as you were limited to their custom syntax kinda similar to BBCode, and I think there was a character limit on how big the body of your site could be.

Their clipart options were kind of great though.

I wish I could find any of my old Expage sites on the Wayback Machine.


So as to not totally derail this topic with Expage memories, another confession of mine is that I hate ranch dressing.

Most of what I visualise in my head isn't what necessarily comes out in the end. But that is the fun of it too, to see how close or how different we get to our vision. And then, we can try again if we desire. Part of the AI bros argument is wanting perfection. We value it too much. We need to enjoy the process again, playing with traditional materials or seeing which digital brushes make what results by our own brain, hands or body parts (if you cannot use your hands).

...

I wish I had waited a few more minutes before my last reply so I could have seen this one, because I just want to say that this is a beautiful sentiment. I'm definitely guilty of perfectionism whenever I make something, which is probably why I have used AI to help with the portions I can't create myself rather than accept my limitations. I have an image of what I want in my head, and I'm going to make exactly that

Also, to your point about people with physical/mental disabilities adapting to create art in spite of (or inspired by) their limitations, that goes right along with what I said about appreciating the work that goes into practical effects and physical puppets vs CGI, and rejecting "death of the author". To me, knowing how much effort someone put into their work has a huge influence on how much I appreciate it. Artworks entirely generated via AI are just a novelty to me right now because it's amazing (and often hilarious) what software is able to accomplish now.

Yes, that is absolutely better than doing it with an AI tool. If this is going to be the status quo (which I absolutely hope not), imagine if we actually lost the art of drawing in a century's time or so because people didn't want to learn to hold a pen and draft something themselves.

Imagine it's because this is what society wanted now. They didn't want artists, they wanted a fast food version with zero human emotion and input. And because it made money, people wouldn't want to be artistic anymore.

We do have to think of longevity here, cos there are people as I said starting to feel drawing, painting, etc is pointless.

I don't disagree, but what you're describing is a negative feedback loop resulting in an extreme dystopia and is itself a "slippery slope" argument. If all new art is created by AI trained by traditional art, to the point where people eventually lose the skill to create art on their own, then art itself will come to a standstill because AI will have no original art to train on. Eventually all of the permutations that AI can generate from their original input will be exhausted, and even before that it'll lose all novelty as all new art becomes too similar.

As long as humans value novel art, this isn't a situation we need to worry about because the need for something fresh will inspire people to create new art. Using your fast food analogy, just because fast food exists and a lot of people want it and can't afford a gourmet meal, that doesn't mean the culinary arts have ceased to exist. Many people still value a home-cooked meal or food prepared by a master chef.

Of course artists deserve to survive on their art, but is it not time then that we get an economic solution on that as a whole? Why do we have to lock down our cultural products and make them stingy commodified property just so artists can get crumbs for their work by forcing people to pay for it (arguably making artistic expression into an economic act of marketing instead of a passion) when we could have a general basic income that anyone can survive on, without strings attached?

Artists could therefore be paid a living wage without restricting their works with copyright and getting up in arms about perceived slights. I believe artists are way too protective over their own works and folks "stealing" them when in reality it should be a general cultural good, a public domain to partake in. The notion that our human lives are here to produce and treat society as a potential customer base is dystopian.

I say that as a musician and a programmer who has published all their works under free copyleft licenses out of conviction.
I think it is unfair to characterize anyone who enjoys the idea of algorithm assisted artistry as an evil artist hating tech bro. We don't have to draw battle and property lines over art until everything is proprietary and locked down.

100% agree. We just need to get all of the world leaders to agree to abandon capitalism.
That was a joke, but I absolutely do want this for society. I just think we need to separate what an ideal world looks like from what a practical one is.

You mentioned you're a big Trek fan, so I get where your utopian ideations are rooted in. But the perfect world where we can all do things for our own betterment and the betterment of society instead of what pays us what we need to survive isn't going to come through a hard cutoff. We'll get there gradually, or we'll kill ourselves off. But that's not to say we shouldn't be pushing for the ideal world right now. It's people pushing for it that will lead us away from our inevitable demise. We just need to learn to be happy with limited progress in our own lifetimes so that in the future others can benefit from our effort.


Also, I just wanna say that I'm glad this topic sparked so much conversation and passion. I wish every thread I started could reach two pages in less than an evening.

But there in lies the problem. By using prompts, and knowing to use which specific prompts, all you're asking the tool to do is to try and create an image in the style of someone or something else with nothing of your own creative input in the first place. All the prompts are doing are just pulling from a source that, more than very likely, someone or something did not consent to. It's a complete joke and laughable to think there's "skill" involved in doing that. It's completely disingenuous, especially with the majority of AI tools are using artwork without permission.

Comparing using select prompts to, I don't know, 20+ years of a human brain studying different artists and then coming out with your own unique take that no one else will ever be quite able to replicate because humans will always have their own flair. There is nothing to compare there.

I might be thinking of AI being used in a different way than you. If you're telling the AI "make a comic strip in the style of Garfield about Garfield doing a Garfield," yeah, you're just copying someone's style intentionally. But is that really better than you drawing a comic strip in the style of Garfield about Garfield doing a Garfield with pen and paper? I don't see the AI as being the issue here. It's the person making the image who is committing the copyright/trademark infringement.

When I've used AI to generate an image, I'm typically using it to generate pieces of a larger artwork. Something like metal ductwork running at a specific angle, or the texture of skin under certain lighting conditions. Anything that I suck at painting myself despite how long I've been doing it. In this sense, I think it's more like using stock photography in a composite image, but I'm able to generate exactly what I need (with a lot of tweaking) instead of hunting endlessly for something. I guess that threatens the livelihood of stock photographers, though...

But like you said, consent and copyright are absolutely still issues in play. If that metal duct I generated to put in the corner of the background of an image was generated using photos that the original photographer didn't consent to being farmed for algorithm training, that's a huge issue that the creators of these AIs are only now starting to address.

As just a little funny aside, to anyone who thinks AI can't be artists, just look at their inability to render hands! All fledgling artists struggle with hands.

Never heard of Expage before but I searched it now and discovered a tale of early 2000s horse website discrimination  Web stables made on Geocities and Angelfire etc refused to cooperate with the inferior Expage stables, which lead to the founding of the Association for the Support of Expage stables. They wrote (translated by me): "The discrimination and belittling of Expage stables on a MASSIVE scale has begun recently. E.g. compared to Geo-stables. On Expage, you can see stables equally as good as on Geo Therefore, we started this association - for all expage stables. Every Expage stable is welcome to join us and we also wish for it from the bottom of our heart!"

Interesting piece of history...

That is hilarious and I love it. Thank you for sharing

It's so nice how such an obscure piece of internet history was somehow preserved.

You will not frigging believe who misclicked something and erased an entire post again. Also there's been three posts posted while I was rewriting this so sorry if I step on someone's point. ANYWAY.

I know the feeling. I've been trying to reply to this for so long, but keep seeing people reply with great points that I want to address so I have to go back and quote them as well. Welcome to the list

I think I cover a lot of your points in my response to HayleyMulch when talking about consent/copyright and artist livelihoods below, so to keep this post from getting too long I'll just add this:

What we also have is people making "art" in an artist's style not even a day after said artist's death, showing great disrespect.

This is EXACTLY how I feel about "holograms" and digital recreations of dead musicians and actors. I'm really disgusted by it. It's one thing for Rogue One to have a de-aged recreation of Carrie Fisher for a brief scene at the end (she was still alive by the time it was released and presumably authorized the use of her likeness), but replicating Peter Cushing's entire performance for a major role is way too far.

James Earl Jones recently authorized the use of AI to replicate his voice for Darth Vader in future Star Wars projects, which is fine, but that's different from taking the face, voice and acting style of someone who has been dead for a long time and never could have consented to it.

As an illustrator myself, I am very staunchly against AI art. Sadly the majority of those singing the praises of AI art have actually just shown they do not want to pay or support artists.

...

Now writing and photography isn't my lane. I'm speaking here in the realms of digital illustration. I absolutely believe there always should be consent and artists permission given first. If an artist wants to make images specifically to feed an AI program, than great! But I still fear the future it will give for working class artists. Corporations will want to use it as it's cheaper than hiring human labour, and I have genuinely seen people on bird app say "what's the point in being an artist anymore, I won't do it", which is so dystopian. It's horrible to rip away an ability that's innate within all of us.

And I disagree. I think there is zero skill whatever ever needed to just type a few words into a program to play lottery on what you're gonna get as an image.

EDIT: while I did say 'writing and photography isn't my lane', I do believe those creators should have their rights, copyrights and consent given first at all times.

I think that consent is absolutely a must when training any AI, and I don't think the "opt out" method most sites are giving is the right one. People should have to opt IN for their copyrighted works to be used for any reason and people should be able to sue for copyright infringement if something they own is used for any purpose they haven't authorized (although the TOS of most sites that do this are being updated to say you implicitly grant them permission to use your art for AI training, but the most popular AI tools right now were trained by scraping Google images which definitely doesn't include any form of consent).

And I absolutely want artists to be able to make a living off of their art, or else art and culture as a whole will suffer. Like I said above, artists need to be able to make a living off of their work because the skill they possess and the amount of time it takes to create their art means they don't have time to learn another skill or do another job to pay the bills. Artists should be able to make a living as artists.

But I've seen people make the same "zero skill" argument when talking about digital vs analog media. AI is a tool, just like any other, and all new tools are created to make it easier to do the same task that our predecessors had to struggle with. Traditional artists threw their ire at digital artists because Photoshop made it too easy to do the same thing that they were doing, and I see the same thing happening against AI.

I have done a lot of illustration in my life using traditional media, digital media and now supplementing it with AI, and I find there's a lot of skill involved no matter what you're using. Coming up with an AI prompt isn't as easy as you think. If you're just throwing whatever in to see whatever comes out, it's absolutely mindless. But if you're aiming to generate a specific image, you need to know what words and phrases the AI actually responds to and also be able to anticipate the result. It's just like how you need to know how different brushes respond to pressure and how it'll react to the texture and material of the canvas if you want to be able to take the image in your head and express it externally as a painter. And it's also like how you need to know what digital brushes or filters will create the effect you want as a digital artist.

Using the brush tool in Photoshop is really just a bunch of algorithms responding to input, just like an AI. If you looked at the code behind where you touch the screen and what the outcome is on the digital canvas it'll look more like an AI prompt than a painter using an analog brush, because in the end whether you're interacting with AI or Photoshop it's all just a program. The AI is just able to do a lot more with a lot less input.

The hard truth is that its not really up for debate; AI is not going away; its going to get better and it is going to be used by a lot of companies and people who need things created. I think AI will replace the jobs of many print artists, comic book authors, UX designers, graphic designers, video editors, audio engineers and many many more! - Its not nice to be replaceable, but it happens

Yeah, I often wonder how long until a machine can do my job. I have a friend who was actually using OpenAI to generate an extension for Visual Studio Code. Without knowing anything about TypeScript, he tells the AI what he wants the extension to do, then he attempts to run the generated code and sends the errors back to the AI to have it correct them. I should follow up and see how far he got with that.

Is AI created imagery Art though? I don't think so. Good art is about the human soul, it has to say something new - the one thing an AI can never do is say something new, it does not have the ability to interpret true meaning or play on our expectations - the day an AI can do that is the day that the AI will be indistinguishable from a human - anndd we don't live in Blade Runner just yet

Art is a really nebulous concept. I don't think we can definitively say that the artist's intention, soul or creative spirit is necessary when, at the same time, critical theory intentionally disregards these things. That says that the viewer's interpretation is what gives art meaning.

But I, as a "consumer" of art, don't really agree with "death of the artist". I think that the context of the artist's life is absolutely important in interpreting the art, and I definitely value the effort and intention that went into creating it. That's why I have such a love for practical effects, costumes and puppets in movies instead of CGI. It doesn't necessarily make the final product better, but I absolutely appreciate it more (while accepting that other people might not see the same value in it, because it's subjective).

But at the same time, I'm on board with your excitement about the technology. If I find out something was generated by an AI, I find it fascinating by virtue of what a machine was able to accomplish. Whereas if I saw/heard/read the same thing created by a human I'd think it's kinda boring. Like, have you seen those funny scripts where people say "I fed an AI 1000 episodes of ______ and had it generate it's own episode"? Those are often hilarious to read because you can see all of the quirks of the source material that the AI exaggerated. But then, when you find out it's fake and it was actually created by a real person mimicking an AI, it loses its charm.

Which reminds me, that's another use I've found for AI. The way it exaggerates the most notable features of something when it tries to replicate it. Like, if you see an AI generated image of an actor, it'll exaggerate the shape of their nose or their eyes or whatever else is most distinguishing about their appearance. I find that really useful when making my own portraits or whatever. I'll actually copy the AI as a sort of tutor telling me "this is how you make that weird generic blob person look more like Adrian Brody" or something.

It's similar to how I used to want to be a voice actor, so I do a lot of voices. But there are some voices I was never able to mimic, like Christopher Walken or Morgan Freeman, until I heard another impersonator do their impression of them, because they exaggerated what makes their voices distinct, and then I was able to replicate it myself.

white chocolate

I used to loooooove white chocolate, but it's way too sweet for me now. I always feel sick after I eat it

In my opinion, art in general is something that should be common property throughout society as a whole, not an individual commodity that is protected and sold by only one artist. After all, everyone, including artists, is a product of generations before them who influenced and shaped them into who they are now. Without ever having seen a portrait, would I truly have painted a portrait the same way I did now? Probably not. Every piece of art created is really the work of all the generations that came before you, including you. The colors, the tools, the canvas, they are all products of generations upon generations of working class artisans. No piece of art is truly only one person's product.

This is something I think about every time I see someone post their charcoal artwork that looks 100% like a black and white photograph. It's amazing how good some people are. I haven't seen any classical art that compares to the realism of modern artists.

So I wonder, is it the tools that are better, or are present day artists that much more skilled? If they're just using charcoal, I don't think it can be the tools. They must have built on all of the lessons and skills passed down from artists over centuries.

I can only see sad, capitalist reasons for maintaining full control over one's own art: the idea that everything we produce, everything we enjoy needs to be a commodity that makes us money and is one person's property to do with as one see fit. Sure, artists still need to be paid, but I don't think algorithmic artworks will kill manual art any time soon; and even if it does, there are plenty of professions that appear and disappear with technological progress; and it never meant to stop the progress of society.

In an ideal world, artists would create art for the sake of creating art and getting acclaim from their peers, and if someone enjoys playing around with stable diffusion algorithms to create art they view as beautiful, so be it. Life shouldn't be a competition about who put in the most effort or who suffered the most. After all, we all kind of just want to live and float through life with a happy feeling, right?

This is where I kind of disagree with you. While art for art's sake is a great ideal, and something I'd like to see humanity reach someday, it's just not the reality we're in. You have to remember that a lot of classic artists throughout history enjoyed "patronage", where rich people basically gave them enough to live a comfortable life so that they could produce art to their liking. You could think of it as the original "selling out", but everyone needs to make money (for as long as money and scarcity are a thing in the world) and really great art requires so much devotion that it doesn't leave room for anything else to pay the bills. I don't think art as a whole would be as remarkable if we were at the mercy of hobbyists.

But you're definitely not wrong, and even copyright law acknowledges what you're talking about. If whoever created something the first time had indefinite rights over it, so no once could replicate or build on it, we'd completely stagnate. So copyright only lasts the life of the artist so they can benefit from their work, plus however many years Disney lobbies for so that they don't have to share Mickey Mouse with anyone. That's where copyright law kind of falls apart, but at least it seems to be catching up and Disney's most recent attempts to increase the timespan copyright is maintained after the author's death have failed.

I dislike AI art. That being said, I'm willing to tolerate AI art if the prompter is 100% transparent about it and is not using it to get money or... "clout". On the topic of copyright, however, I don't think the prompter should be able to copyright the art that the AI makes because, well, the AI made it. If anyone should be able to copyright it, it's the people that made the AI. The most that a prompter could hope to copyright is their prompts as that's what they made.

Uncopyrightable AI art (partially) solves the problem of AI stepping on the toes of professional artists. People using AI art for projects may be able to use that AI art, sure, but they will lose out on copyright for whatever the AI makes. Having to regenerate the image over and over again may take time, but (and I don't say this with the intention of insult) there isn't much skill in pressing a render button until the image is satisfactory. Giving prompts to an AI is like to giving prompts to a freelancer. The prompter of the AI is more like the client, not the artist.

Your point about client vs artist is pretty good, but I wonder how you feel about photography. The image already exists in front of you, and all you need to do is push a button to capture it. Professional photographers will take dozens or hundreds of photos of the same subject in quick succession and then select the best from the set. They also need to worry about framing, lighting and composition. But are these things really much more of a contribution on the artist's part than refining a prompt iteratively and pressing "generate" over and over before selecting the best result? Even if an amateur takes a quick snap of the landscape without any thought behind it, they own the copyright so the skill behind capturing the image isn't what distinguishes them as the creator/artist.

AI is a really controversial topic right now, with artists protesting against their works being used to train algorithms without their consent, or AI art being used (and in at least one case winning an award) instead of hand-crafted art which threatens the livelihood of professional artists.

Then on the other side, there are people being denied copyright because they used AI generators for all or part of a work they lay claim to. I personally think that is a bit extreme because it does take effort (and a lot of patience) to craft the ideal prompt for what you want to generate, and there is a lot of editing and decision making when deciding what to include in the final piece for compiled works. AI is a tool that people work with to generate the desired outcome. You could argue that it's like using digital art programs like Photoshop. Yeah, one takes a lot more effort to create something than the other, but they're both tools so I'm worried about where people draw the line between tool and creator.

In the case of the guy who generated an entire children's book using AI for both the text and images, I think I'd probably consider him more of an editor than an author. But it would depend on whether he just went with whatever the AI gave him and threw what worked together, or if he spent a lot of time regenerating to get the result that he wanted and then touching it up after. I can sympathize with that, because I often have an image in my head that I want to get out, but still, after decades, don't have the skill to realize it.

That said, I think there are legitimate uses for AI that don't harm other artists or detract from the artist's claim that they created the final piece.

I have used AI in the last couple of years to help me when I'm stuck with writing. I give it what I have up to where I'm stuck, and then let it generate the next passage to get me over that hurdle. It usually involves clicking "regenerate" over and over until I'm happy with the general result, and then I do a lot of editing to make it fit and put it in my own words.

I also have used AI text-to-image generators to create pieces of a larger artwork that I'm struggling with. Again, it requires a lot of trial and error to get the prompt and result to look generally right, and then takes a ton of editing after to adjust the lighting and painting over the seams to make it fit into the overall image.

But maybe you think even those two examples are going too far. Let me know what you think.

Either way, I don't think there's any stopping algorithmically generated art. Even in our utopic scifi depictions like Star Trek, they have the holodeck which can generate an entire virtual setting and characters from a voice prompt. But Star Trek also explored the ethical ramifications of this in a couple episodes, like in TNG when Lt. Barclay created virtual versions of his crewmates and in DS9 when Quark is asked to create a digital likeness of Major Kira for some rich guy's sick pleasures in the holosuite.

I recently started watching Doom Patrol, and this song was in the first episode so now I'm listening to it all day.

Actually, a lot of great songs were in the first season, like "Lazarus" by David Bowie.