I’ve been pretty quiet over the last few months, watching curiously as the debate against AI text to image generators have become more controversial. As a fine artist and designer, when I first starting playing with midjourney, I was impressed with what could soon be possible. Version four has already defeated the main claim that AI art will never be as good as real art. It’s often much better than what you could hire an illustrator to do for you.
But there are other arguments. I’ve tried, in comments and FB groups – even when being attacked and accused of being anti-artist and pro-AI – to be polite and reasonable, but it’s no use, because there’s no benefit to arguing for robots over artists.
As Chuck Wending says,
I’m going to side with the artists. (Spoiler: you should always side with the artists.)
And he’s right! I agree, and I also support artists. I agree with everything he said in this blog post. How “artists are like dinosaurs getting mulched into oil to fuel this thing.” All that is true: AI art will mostly benefit the techbros, and harm the artists.
But he ends with this comment: “I suspect (or at least, hope) this AI art thing burns out.”
And that’s the one place we disagree. Because I don’t see it burning out, or going away. It might be bad (yes) but it is here (today).
Artists were aghast about a partnership deal between microsoft and dalle2 to create a stock photo site. Now DeviantArt has jumped on board with their own AI tool based on the same software artists universally hate. I expect in the next year, AI image generation and even text generation will become far better than we expect, and it’ll also become absolutely ubiquitous. Just today, Adobe Stock officially recognized AI art… so all the stock photo sites will start having AI stuff for sale, which means you can’t avoid it, however you feel about it.
And Canva just launched its “Text to Image” generator; Microsoft launched its Designer days later, an online graphic design tool with the option to generate images from texts (rather than searching for stock photos).
Graphic design software made it easy to import/buy images from the big stock photo sites; but that meant they got less profit, and you got something less unique. Of course they’d prefer you use images that cost them much, much less than the stock photo sites.
As artists and writers, we can be angry about it, but we should also make sure we figure out how it’s going to be used against us, how it’s going to hurt us, and find ways to continue doing what we love without being replaced by robots.
“Hope” might be the best, only strategy, but it’s unproductive. Here at Creativindie, we find specific, practical ways to harness creativity into objective value. So I’m taking the devil’s advocate position, even though it’s unpopular. I’m not cheering for the techbros. But I hope I’m helping artists deal with he frustration and anxiety, and preparing them to deal with changes that humanity has never before dealt with.
There’s really no point in discussing the merits, limitations and scandalous “theft” of human artwork, but I plan to do it here anyway for those who have reservations, or have been influenced by memes. I hope to convey that this isn’t my firm opinion, I’m not “pro-AI” – but I am aware of what it can do, and I recognize that it is creating an enormous shift in how we get things done, it’s going to have a tremendous influence on all kinds of businesses.
As an author, I need to be aware of it and how to recognize it, so I can spot AI art on stock photo sites or avoid paying way too much for an unscrupulous cover designer who claims it’s custom human art when it isn’t. As a writer, I need to pay attention to the fact that it can improve my writing, save tons of money on editing, and some of my competitors are already on board. I don’t have to use it in my business, at all – but I need to consider it, and keep a cool head.
I’m pleased to have studied art history in Florence and moral philosophy in Malta; while also being a fine artist and a working graphic designer for many years… which doesn’t mean I’m right about anything or that you should listen to me, but I also promise a balanced, nuanced investigation into some potential humanity-reducing technological developments we can no longer ignore.
PS. most people are taking moral stances against AI art right now, and/or declaring that they will never use it and their art is “real” art. BUT, a lot of these people also depend on stock photo resources, and it’s going to become very tricky to deliberately keep AI tools out of your workflow, as they start getting added into all the tools you’re already using.
Maybe I’m being pedantic, but in Chuck Wendig’s post, he shared a viral tiktok video against AI art: and the creator’s profile link was beacons.ai – a platform-building software built on AI, so you can get set up faster, with less work. Obviously, not a big deal, because AI in general might be fine, while AI art might not. But it is slightly ironic, and you could make the same arguments that AI website builders have “stolen” everything from coders and developers, making them obsolete and putting them out of work. So we’ll need to have a discussion about what kind of creative work deserves protecting, and how to decide…
The arguments against AI art
People think because I talk about emerging technology (“augmented creativity”) I’m pro-AI… but the two comments above are more in line with my feelings. Humanity was already doomed, and there’s a slim chance AI might help us get our sh*t together. I’d prefer not to talk about it at all, because it’s so polarizing, and it’s a little bit like arguing about an incoming meteor: yes it will destroy everything. There’s some brutal upheaval coming.
But there’s also a lot of bullying and disinformation going around, two things I can’t abide. Artists are telling me they feel unsupported; but they are also cursing me with dire threats and violence.
I’ve yet to be proven wrong that AI is actually doing anything illegal, by browsing millions of images and learning to draw in similar styles. Art styles can’t be copyrighted. It’s true AI has learned (and could only have learned) from human creativity. Mermaids don’t exist in reality, so all “mermaid art” robots create is based on artwork, not photographs. It’s understandable that artists feel cheated and violated.
It makes sense for them to protest. But calling it theft doesn’t make it actually illegal. With legal issues murky at best and unlikely to ever be properly restructured to ban AI artwork entirely, that leaves only the moral issues, which most content creators understand. So they say things like this:
“They will argue that it’s legal, but we can all plainly see that it is not just.”
So, this is an issue of justice. Basically, it’s not fair. I’m not arguing that AI art is cool, or good. I’m just telling you that it’s here. But that’s still a step too far.
After reading hundreds of comments, I’ve boiled it down to three issues.
Real art has a soul. AI can never replace human passion.
This feels true, but it’s also something amateur writers often say to justify their work, despite being untalented. They equate raw passion and joy in creation with quality of output, and it rarely is. Passion and joy in the process can help you to keep going, until you learn enough skill to be marketable, and able to create quality of work people want to buy, rather than staying a hobbyist.
It’s the same argument used against successful authors or artists who are writing or creating commercial stuff on purpose, making money with their art. There’s an idea that if they aren’t suffering for it, if it doesn’t cost them, or if they aren’t enjoying it (putting joy before money, the literal definition of amateur, one who does for love) then they aren’t “real” artists.
Think on that for a moment, because this is a deep core belief throughout creative communities: that we all desperately want to be good enough to be adored, but also want to just do what we love, and often have to find a balance in between our unpaid creative work that only we love and soulless, banal work we hate that clients will pay for.
At the end of the day, however, how the artist feels about the work, how long it took them, how happy or unhappy it made them, does not usually translate to the work itself. The clients pay what they think the art is worth to them.
When impressionism first became popular, it nearly destroyed the art world. They were ridiculed as unskilled amateurs, everybody carrying around easels outside, catching the light without skill or understanding of the craft… art was a tight community that supported its own. Until it became fashionable and the rules changed.
It’s the same argument Van Gogh used against photography: pictures can’t capture a human soul like an artist can, because it’s just a machine – so he started to focus more on portraiture in part because he was afraid new technologies would threaten his own value.
Interestingly, the various crafts of artisanship were usually looked down on as simple, repetitive, low-brow labor that could be learned and studied without much intellect. But it always focused on skill rather than creative genius or inspiration. The Greeks of course considered many aspects of the muses, but separated the technology or technique of making things from the “divine fury” of guided, almost transcendental epiphany.
But our current passionate beliefs about creative inspiration are nearly all modern, based on romantic ideologies that were at their own time, absolutely shattering to the centuries of careful wisdom and study that had gone before. Poetry for example, had crystalized into specific forms and schools and rules for the “perfect” and “best” experience; when romantic and modernist writers began to break these rules and get popular anyway, they were reviled by the establishment, like the impressionists.
Skill must be earned. If you can copy a famous painter, you earned it.
Art has value in proportion to its cost (resources + time invested + skillset). Artists have argued their prices are justified, because they went to art and design school; they have bought the tools and workshops; this art stuff is hard, not free, not cheap, because the investment is too great.
But! The flip side is that art is worth whatever consumers are willing to pay; and in a time of infinite content, they aren’t willing to pay anything unless they absolutely have to. They don’t need to buy art if they can print out posters or right-click save images they find online. But they can’t use it commercially (few want to).
I would argue the value of art is how many people respond to it, like it, for it’s immediate aesthetic value. Is it arresting. Is it worth their time and attention? Do they stop their scroll? I would argue that the art I’m already getting from midjourney is as good or better than some of the best, top quality artworks I’ve seen in my lifetime. And of course it is: it’s processing the history of human creativity and taking it all one step further.
Why this argument matters: people are saying, yes it’s not illegal but it should be, because humans cannot compete with computers and this always has been and should continue to be a purely human space. Nobody is denying that all artists copy, that all artists process a lifetime of images and styles, and sometimes deliberately copy to improve their skills. But that’s not IP theft, that’s just how creativity works.
Technically, collage is legal, even though there have been some outrages when an artist is found collaging scraps together from other people’s work.
And even though the robots are doing the exact same thing (using images as reference, not copying any one particular thing) it can’t be allowed to continue – because it threatens a sacred human endeavor.
So it’s about banning robots from encroaching on artists, which is uniquely different from all other things (like robots taking over the jobs of amazon workers, which is happening as we speak) because art is special, and must be protected. It’s not just a skill-based occupation. It’s more than the sum of its parts.
So, if humans are talented, after having practiced for thousands of hours, the fact that robots can do arguably the same thing (a sequence of steps that creates an artwork), is not a fair comparison because humans have a soul and robots don’t. Humans paid a price to earn their skill and robots have not: they’ve stolen that burden from artists.
Humans are entitled to creative joy
The final argument is, even if AI art isn’t illegal, it’s definitely bad. I’ve seen creators talk about how this is definitely the end of humanity in general: if art-making has been the quintessential thing since the beginning of civilization (and long before), this is an “unmaking” of ourselves as humans.
What will we do for hope, joy and inspiration when art is no longer an enjoyable hobby or pursuit, when there are no benefits for learning and skill and practice, when art is no longer a difficult challenge?
Paraphrasing one youtuber,
This is artmaking, this is joy: don’t take joy from humans and give it to unfeeling robots. Receive artmaking for those… dumbly bestow it on unfeeling dumb being who can’t enjoy fruits of labor.
That’s the whole argument. It’s not illegal, but it should be… why? Because? New innovation will put people out of work. It’s anti-humanist, amoral or immoral, anti-artist and creativity, the end of the world. And they are mad especially when people tell them it isn’t a big deal and artists can use it too… I don’t believe that.
The founder of stable diffusion, I believe, said this:
“I won’t allow us to build tools to replace artists. It’s simple as that because it matters to me. Let’s be clear, these tools require a skilled artist.”
But I think this is a replacement for artists, both figuratively and literally, it’s an enormous, gutting change, for the history of human civilization. We’ve never been here before, at all.
Someone on my youtube commented:
this means art and creation itself could be a phase of evolution that comes and goes that was never really our main source of identity…
And yes, this is an identity crisis for the whole human race.
Someone else said,
This current wave of interest in AI is a novelty, the pet rock or Rubick’s Cube of creativity. It’s a party gag-level toy. Anyone remember “Mad Libs”? That’s what this AI writing reminds me of.
Those are basically the two choices: should we ignore it and hope it goes away; or recognize that this could be a humanity-shattering event; the singularity. We just had fighter jet pilots simulate a fight with an AI fighter plane. SanFrancisco just allowed robot dogs with deadly force (explosives). None of this is good. But it’s not a party-gag.
Quick Wrap up (visual guide)
Here’s a meme I’ve seen floating around.
That’s misleading, because AI does not make a collage with images like that. It creates a totally new thing. The main point is that, because robots can’t be inspired (because they aren’t human), even though they are doing the same thing, it is not OK (because they are too good and too fast). The human brain can look at a thousand pictures and create something new, but robots cannot (even if they technical can), because don’t deserve it.
However YES there’s some weirdness about how the model was trained and it’s fine to ask questions and even feel outraged… as long as you don’t bully others for using something that they enjoy (and you might say “but they need to be educated!” and maybe you’re right, but to what end? Maybe you’ll convince all the content creators to boycott them? The content creators don’t fund the design industry: paying clients do. Clients still need deliverables.
This meme is more accurate:
Yes it’s POSSIBLE, even if you don’t use any famous people, that AI will search for “epic portraits” and find the most popular, which will probably be celebrities, and use that to guide its decision, so you might get something vaguely george clooney looking. That’s a consideration, and it’s problematic.
You can also add specific people and get stuff like tom holland in underwear. Scary, deep fake stuff. But you don’t have to violate people like that (sorry Tom!) PS. those boudoir pics were made on midjourney2 I think, so the quality isn’t good… compare it to some more recent stuff down below. These are just a few rough examples… most impressive are the historical portraits (girl looking at a camera, polaroid, [year])
“But I can see the copyright logo!” – yes, midjourney looks at all the art, and then guesses at what I want it to make. Since maybe 50% of this kind of artwork has a copyright or artist’s signature, AI vaguely tries to reproduce that. It isn’t anyone’s specific logo (though sometimes it is… big stock photo sites have millions and millions of images, so it’s weighted… and not impossible to get results with a shutterstock logo). I know, that should set off alarm bells and be the nail in the coffin against AI art.
But again – not illegal to browse free photos and use them as a basis to inspire yourself to draw something. A human artist would just remove the logo, the AI hasn’t learned to yet. It’s weird, but I don’t need to pay for photos unless I actually need to use the logo-free version commercially; I can look all I want without paying. The argument is that AI should/must get permission/pay to “look” at all of those billions of images. And who knows, maybe they should – I’m not a lawyer – but here’s what I can tell you: they aren’t going to, and this has already happened.
Also it garbles text… but I bet in a year, you could put in “text you want to use” and “fonts to use” and it will make it happen.
Incidentally… remember that meme I mentioned earlier about how AI ART IS THEFT? It’s part of a big slide show about why you shouldn’t use Lensa because AI art is bad. It’s poorly designed but effective propaganda. It fires people up so they love to share and support the message (which is fine).
BUT at least one of the people who started resharing this meme, took the original slides from Gen Ishihara, which had 2.4K shares, manually edited the slides to add his own name in the acknowledgments, and now has 35K shares.
This is the post that everyone is sharing to boldly disavow AI art as theft; and the entire post has been stolen and edited without permission from the creator. Gen Ishihara had to update her post to say this:
Edit: I am fine with people reposting but PLEASE DO NOT CHANGE THE INFO ON THE SLIDES i cant believe i have to say that but wow
It’s trivial… maybe. Or maybe it’s the entire thing. Be open to the potential that rallying people around a cause is great for getting viral visibility, even when you’re doing the thing your post warns against. That even if well-intended, you might be manipulated with untrue statements about how all of this stuff works.
I’ve even seen cover designers take this same stolen meme, edit it to focus on whatever new AI threat looms, and continue pushing the narrative that AI art is theft even as they steal from other artists.
There’s also a case to be made that “human creativity” has always been about stealing. That’s the literal definitions in most nonfiction books about creativity in the past twenty years: that humans have the unique ability to combine things in new ways in their clever brains.
“To be a master of metaphor,” Aristotle wrote in the Poetics, “is the greatest thing by far. It is the one thing that cannot be learnt from others, and it is also a sign of genius.”
You can’t really find any popular artist in history that wasn’t a product of his/her time; which meant, copying what was popular and using the best tools to create more content. Some more more egregious than others, like Picasso, whose friends would not let him visit because they knew he would still their ideas. Stealing (being inspired by) art styles has been a normal process for most artists, even though the history of their plagiarism is ignored in favor of their eventual unique synthesis.
Philosophically, it’s an interesting battle. Most people don’t need to make art; they do it because they enjoy it. Most “regular” people don’t have much use for AI art… except for making cool new profile pics. Lensa is going viral quickly, and the conversation around AI has exploded at the same time. Most people have seen the memes and reservations. So it’s a personal choice: does my desire to look pretty on social media outweigh my compassion for those it might be hurting? I’m not optimistic, but if making art is quintessentially human, maybe so is vanity and prestige (to make and share a fun new thing, while posting flatting selfies… after all, isn’t that what almost all art has been since the beginning – a way to capture our personal essence or expression while we can?)
As my friend Jenna Soard says, you should avoid any specific artist’s work or name in your prompts, and AI images should be labelled as such and not disguised as “real art.”
There are ways to “Ethically” use the knowledge of 500 Billion questionably sourced images. And there are ways to steal on them. Don’t be an asshole.
Another FB friend of mine, just upload 150+ new selfie photos made with Lensa.
AI art vs. stock photography
This one is a personal gripe… on two counts.
First, the majority of working artists aren’t really making art. They are usually selling commodities – getting paid for a specific result. It’s often technical photobashing and manipulating stock images in photoshop, maybe with some hand-drawn elements. Not exactly the same as full custom illustration. The work they are doing, while certainly creative, isn’t exactly inspired. They will probably try to make high quality works based on similar products.
Most artists don’t get paid at all. Those who do are *often* making products. I’m not saying we should take that away from them, but it’s sometimes tedious and relatively easy to do basic things, and robots could do it faster (people have been using bought photoshop actions for years, that just basically add a filter.) Same with fonts or text styles. It takes a creative eye to put it all together, but they are mashing together stock resources.
Second: even when they use safe, stock photography, and even when they swap faces and add a ton of details and make everything much better… often the amazing art will appear too similar to anybody else using the same stock photo, because they aren’t unique, which is why authors are sometimes accused of copying each other. AI art allows for unique art.
Designer reactions: designers are in a tough spot right now, because they can either fight against the tide and become unpopular, or they can embrace the change and also become unpopular… but I expect in a few months, things will die down, and I’m starting to hear some optimism.
If AI is already on Canva, on stock photo sites, everywhere, it’s going to become mainstream enough that people (paying clients) don’t have a problem with it. And it’s really pretty incredible, because you can make great stuff easily. But MOST authors/clients/whatever, probably aren’t going to mess around or get their hands dirty with AI tools. It’ll take them years to catch on.
AI is great for making very specific things, like special armor or a background landscape. It will make it easier for designers to make better things (this is specifically about, for example, fantasy book covers, which have always been difficult to pull off with stock photo or without hiring an illustrator). It’s possible, that it just becomes an obscure tool designers use to serve their clients.
A shift in the dialogue…
Currently another newish argument is blowing up in the indie publishing world. Some big authors have started using AI art for noncommercial character portraits. They look fantastic and are just made for fun. Some artist are calling them out, basically saying that anything made by AI is stolen/pirated.
They feel betrayed by authors, who are happy to “pirate” real artwork.
And while not illegal, if a certain artist was using specific prompts or art images to generate new images, that’s problematic. But you can make great AI fantasy without doing this.
You can also, however, in a way that wasn’t possible before (with midjourney4), reverse engineer anybody’s art to a surprising degree. So I could take 5 photos from somebody else’s art and make something similar. It’s a lot faster and easier than experimenting with text prompts.
It’s still not technically, actually stealing, but it’s really really close. But just because it’s possible doesn’t mean everybody is doing it; and just because AI was trained on real art doesn’t make using AI stealing from artists (some people think it does – exactly this! – but that’s a moral judgment, not a legal one).
Using AI artwork (trained on art, but creates new art) is not the same as actual piracy (people downloading books for free without using them). The mediums don’t compare well, because a work of art is consumed in a day so people pay for limited access or unique paintings, that they can own, display or use (for example in exclusive book cover art).
A book is often a longer experience, so people are expected to pay, like the price of going to see a movie (or the Netflix subscription model, like Kindle). But people often have to pay much more for custom, exclusive artwork, with the idea that it’s only bought and sold once, and you have to pay for the artist’s time; as opposed to a book, which is sold cheap but can be resold infinite number of times.
Interestingly, for the first time, some big authors are starting to stand out and defend each other against increasingly hostile artists who want to burn everything to the ground. They have a point, and I feel for them; they have no legal standing, so they bully and intimidate, on moral grounds… which they are entitled to do. And for awhile, it worked. Authors were scared of getting called out, of AI art in general, determined to be on the “right side” of the indignation and swearing to support artists.
But that’s shifting really fast. Partly because well-meaning, and well-liked authors are discovering the joys of AI as a fun tool, partly because people are tired of the drama and controversy. And most people are still talking about the versions that existed months ago, when the tools are already 300% better than they were.
The common refrain is that authors shouldn’t support AI writing because AI writing tools will come for their jobs next. And that’s true… but a lot of authors are already using them. They enjoy using them. The violations (if any) on style, theme or writing quality is much less noticeable and never actually plagiarism.
AI writing tools for fiction
A diversion… my audience is mostly authors and writers, but you can skip this stuff and get back to the AI art controversy stuff down below…
People keep saying, I should be more understanding as a writer, because what will happen when AI writing tools can write books? Well, we’re already here, people just don’t realize it yet.
The AI writing tools just help generate content. They were useful a few months about. They are insanely powerful right now (only this week!). By next year, I do think I’ll be able to write an entire (good) book in (my) writing style, that will be nearly indistinguishable from my previous novels, and I think it could take a few days. Which would literally change my life, because I have so many stories I want to tell. (Here’s a sample of AI writing tools finishing my chapter for me.)
BUT – we’re going to be having this conversation all over again soon, with editors and proofreaders; when AI tools are way better than grammarly or even most developmental editors. When you can put in a messy rough draft and it will smooth and polish everything and make it better. When non-native speakers can tell the stories they want and have it cleaned up (the entire manuscript!) for $20 instead of $2000 in developmental editing fees.
It opens the playing field. It’s great for everyone, except the people who spend years developing their skills and businesses. Those people will be the most scared and complain the loudest. Or, they might not. I myself and many authors I know – it’s almost a definitive quality of writers in general – love to have written but hate writing. We struggle with procrastination and motivation, of making our writing good enough through countless revisions. Writing can be fun, and it can also be torture.
AI tools can and will, very soon, make everything so much easier, so you can publish more, better stories… but so can everyone else. I said in a podcast people could use AI to make full covers soon. People think writing a full book with AI is a pipedream… but before they realize it’s possible, some other authors will be doing it well and reaping the benefits.
Adverse reactions: the discussion around AI art is heated and crossing over into AI writing tools. Some authors are already using them to write better, faster. But a lot of others say things like this:
“You should boycott it if you want human authors to have a future.”
“‘Playing with it’ means helping to train it…When it’s fully trained you will have helped put yourself out of a job.”
“Skynet is coming. Don’t say you haven’t been warned. Writers and other artists should be fighting tooth and nail against AI stuff ever, ever, ever being viable. You really want to join the ranks of those being pushed out of jobs by robots? How will we be able to earn a living in the future if computers and robots take over all the jobs?”
I pulled these from a writer’s group, not to denigrate, but because they are strongly worded and popular opinions about AI writing (based on likes and upvotes). There was also a bunch of stuff about how AI writing will never be as good as human, because it doesn’t have a soul, etc – but it is already good enough.
PS. I’ve done stuff like this before: reposted other people’s content as an example, and I’ve gotten in big trouble for it, so I don’t usually do it. But, it’s not illegal, especially since I didn’t cite the source, and even if I had… legally I would be protected. But examples are great and otherwise it feels like I’m just talking to myself. I wanted to listen and really hear the opposing point of view so I could craft a measured response.
Will it take away the meager earnings of a hard-working indie author? Absolutely! Unless they use AI to finish their books faster, and create more content. Nobody is stopping them. They just want to hold everyone back so nobody else cheats and gets in front of them, without skill or hard work.
It’s not fair.
That is probably true, no matter how you see it.
But, the tools are here, they exist, they are
good effective, and some people will definitely be using them to write books soon. It could be you… but it doesn’t have to be.
It could be me… and you don’t get to have any opinion about how I create my work. If my readers love the stories I “write” – will they be disillusioned or feel cheated if they find out I got help from an AI cowriter? Doubtful, but maybe so. Maybe I need to clearly communicate my writing process if AI gets involved (though, people have used uncredited ghostwriters and editors, for a long time; and I doubt many authors credit grammarly or prowritingaid). Still this might be different. I’m not necessarily opposed to labelling all AI work clearly. But I wouldn’t release a book I wasn’t 100% proud of, and AI will probably allow me to improve my current skill set.
Maybe it gives everyone a 25% boost, but from wherever they are currently standing from. With my experience, I could use AI to write a great book, but not everybody could (yet. maybe that’ll change in a year).
AI tools for creative inspiration
I just made a fun writing prompt generator (ok, I didn’t make it… I saw someone share it and made an offer because it’s so cool) – but I’m blown away by GPT-3 davinci’s brain-storming powers. I can have it make a list of names, titles, generate a dozen specific metaphors, answer ANY question. I think I’ll start using it for everything, because it’s way faster than google results and it actually gives me what I want.
For example, when you just need to look up a fact, rather than wading through ads and a dozen results before finding the answer, the AI tool will just give it to you instantly. But it’s also so, so great for brainstorming anything; or writing scenes, scene description, even revising your text to sound better and more polished. It’s hard to explain, but once you start using it, it’s just so much easier to get some good content out. I can even feed it a few paragraphs and let it continue the story, sometimes in surprising ways. If you’re stuck, on a plot twist and can’t figure out what needs to happen, AI can probably resolve it for you. Just prompt it like “protagonist is stuck on a boat with a bomb and an unconscious sibling and can’t swim… how do they both survive?”
AI text to voice tools
Another big arena for indie authors is using voice to text to make audiobooks. Of course narrators are upset by this. And the technology is not quite there yet – but very nearly. Better than those crappy tiktok AI voices. Most authors can’t spend thousands to hire a narrator, because their books aren’t profitable, which gives a big advantage to people with more budget. But AI text to voice narration tools will let you finish an audiobook for $20 which sounds pretty darn good… enough to put up on youtube or something.
Yes it’s a dystopian vision, if in a year or two I can write, design, publish, narrate a book using AI tool for about $50 in 3 and start selling it online, the robots will take over and we will all be miserable (but have a lot of content to consume).
But art has always been the privilege of the wealthy. Only recently, with the internet, and now with AI tools, have new people been allowed to enjoy the benefits of creativity, and share their work to millions, without (investing years to develop) talent and skill required to succeed.
These were some notes I put together for a podcast interview… this is a bit rough. You should probably just stop reading…
My feelings on AI art:
I think we should all have a healthy amount of fear, because this isn’t nothing. This is, basically in the history of humanity, we’ve been dreaming about this point, when our technology can replace our labor and make humans obsolete. We never thought it would happen for the creative arts, writing, artwork, but it has. I mean, this year, after centuries of speculative fiction writing about it. This is a very big deal for everyone, it’s going to change a lot of things, not just for writers and authors.
Coding, for example: people are taking coding courses and getting hired for a lot of money… in a few years coding might be something that’s automated to AI, they’re already working on it. Starbucks drinks automation, food services, Amazon package delivery, UPS, food delivery… automated. This is, really close to happening. It’s probably about 50% of the workforce. So what happens to everybody after they don’t need to work? Art and writing, creativity – but when the robots can already do that better, then what’s the point?
Whether or not you making a living designing or writing, you’re bound to feel depressed about all of it. So the fear is normal, natural. My PhD was in literature so I have a pretty good understanding of classic books, how and why they were written, why they became important and why they matter. But I also work with writers so I know about the struggle, the effort, the work, the satisfaction… of working hard and getting better, of trying to be the best or creating the best work you can.
I was a fine artist, surreal oil paintings, in Italy for awhile, I spent two months drawing the same picture of a foot. Not many drawings, one drawing, until it was perfect. I had to learn on one thing, polishing it until it was good enough. Most people will hire out the skill, on fiverr, character art concepts, for example, full page illustrations, book cover design. But now I can make something much better and more interesting than anything I could have done, but also better than anyone else could have done.
Other than a handful of really famous artists, the vast majority of people are mediocre artists with some talent but nothing exceptional or new. One of the huge advantages of AI art is you can make so much content so quickly. So instead of competing against 1000 artists with one picture, I can make 1000 pictures and have a huge portfolio or gallery. I can post more often, and the work is actually pretty incredible.
Dalle2 vs Midjourney AI:
Text to image tools let you add a prompt of what you want. They started to go public in the summer of 2022; by the end of 2022 they’re about 400% better. It was really tricky to get complex scenes, and they still have trouble with detail (you can get a stunning, photorealistic face, but if you have a big complex scene, the face might be blurry). It’s hard to get a specific character, holding something, standing a certain way, with a certain background. But you can generate multiple pieces and photoshop them together, instead of spending hours and hours searching for stock photos and having to create a whole scene.
Midjourney is still currently BETTER at art that looks good, but I think Dalle2 is smarter, in terms of recognizing the elements of an image and being able to reproduce complex poses (couples embracing, etc). It’s hard to say which will be better because they’re both already decent, but I expect Dalle2’s quality to improve shortly and then might surpass midjourney.
I’ll spend a few hours playing around with keywords, until they get close enough to something cool. Then I’ll spend days sometimes, just respinning content. I made thousands of images to get a handful of truly amazing, exceptional images I can actually use in my marketing; that captivate the essence of the genre, with emotion (which I may not use, until the heat has died down).
Best image prompts:
There are many, but you can generate good art without using any specific artist’s style or prompts like ‘trending on artstation’. That doesn’t mean it’s fine, because people are upset that the data was trained on their art without permission at all, which again… is not stealing (in the legal definition) but they will absolutely tell you that it is anyway. Still I’d be careful with impersonal prompts, avoid artist names, and do something less specific like this:
beautiful young fantasy princess, light elven armor, short skirt, legs, tank top, soft lighting, fierce expression, pet iguana dragon, whimsical art, fighting a demon, dark fantasy gothic, cyberpunk futuristic dystopian, character in full growth portrait. A very beautiful, wide shot, wide background, insanely detailed and intricate, concept art
As the tools get smarter, you need less detailed prompts to get good results. I’d be all for a general ban of all artist’s names in AI art; I’d even be all for a complete purge and redo, where the models are only trained on willing material it has purchased: or a specific “style” charge where artists can submit their portfolio and get compensated if someone chooses their styles. I’d even be OK with a complete ban on AI art until all this stuff gets resolved. Sign me up!
Someone smarter than me will probably build something like this soon, (as DeviantArt are kind of trying to, I think).
This is Spotify, kind of…
There’s a Netflix special on Spotify and it was interesting, because at the time, pirating music was seen as a brave, daring, act of rebellion – free music for everyone! The founder of Spotify desperately wanted to compete with torrent sites that everybody was already using. But why would they pay for something that was free? Because piracy is illegal; but Spotify found a way to charge for unique features. People will pay for convenience. This let them split profits (and get permission) with the real music industry, but it was a desperate fight.
But the conversation at first was one of violent rhetoric: screw these big music companies! Art should be free! Let’s take down the bastards! Later, it changed to become artist-focused (Spotify is making all the money, but barely paying artists anything).
Technology is usually about stealing godlike-powers from those in power and giving it to the masses; this is Prometheus stealing fire or even Even daringly biting the apple. It’s forbidden, but it’s popular. You can’t tell people they don’t deserve/can’t have what they desire, especially when you put it right in front of them and force you to resist the temptation.
But it’s kind of a non-issue, because people care about convenience more than artists (everyone knows Spotify doesn’t pay musicians well, but it’s the only game in town… and most people use itunes or spotify because it’s enjoyable and easy… even if they have to pay a subscription fee).
The new AI tools *might* find a way to compensate artists somehow, but that’s unlikely. I do think they’ll start banning famous artists from search queries or prompts, at least if they complain loud enough, but it’s impossible to do for everybody, and it can’t unlearn what it’s already learned.
The TikTok comparison: a fresh angle recently, is how authors are up in arms over AI art, but *many* authors are also using a personal tiktok account with commercial music, to promote their books…
AI art controversy:
It’s not illegal to look at work before you make your own. It’s legal for humans, this is the basics of most copyright law. I can be inspired by; I can look at a lot of media content and make my own thing. This is what AI does. It’s sketchy they do it without permission maybe, but I don’t need permission to browse famous art and be inspired, I don’t need to ask permission to read a bunch of books and be inspired; in fact I can’t help not to be inspired. I can’t unlearn all the books I’ve ever read. To ask a robot to forget it saw a bunch of stuff, that’s not practical. Dystopian, like you could rent a movie but would have to have your memory purged.
But for using specific keywords, for famous living authors, I think they can make a strong legal case and prevent their name being used as a keyword, I think that’s probably happening or will happen, so you can do a style but not a specific artist.
AI copyright stuff:
As I understand it, you can’t copyright AI work you created. For some people, that’s going to be enough not to use them, because they can’t ever control their content. If I post my pics and cover design, someone could just steal it and use it. That’s probably going to happen anyway, and even if someone does, I won’t go after them legally because it would cost way too much.
ISSUES: I’ve seen people post AI art with their own copyright so people can’t use it; which is effective, but misleading since they can’t actually copyright it. And I’ve seen people selling AI art with ‘commercial use rights’ which they can’t legally do.
But I don’t think that’s a big concern, for me personally, I want to sell as many books as possible. If you put the images together and added some story content, it makes sense you could copyright that new book or product as transformative but it’ll take a few years if ever to straighten the laws out. Art platforms are banning it right now, which is probably good; their market is artists, they can’t afford to piss off artists. UPDATE: no, they are totally not.
Stock photos are a little different. They might be banning now because the copyright is questionable and they don’t want to get in trouble, but there are *already* stock photo sites where you can just describe what you want and get images, so instead of a search feature it will be a create feature. This is going to hit stock photo sites really hard, it might even put them out of business. Why would I want stock photo site that other people have seen, on my blog or website or marketing, when I could create something brand new and unique? These are huge businesses, it’s going to affect things in a big way. What about photoshop: people use it to blend images and layers together to make art. If you can make better art without it, why would you need the program? It was a tool that let designers charge money for their creative content; consumers can now get better stuff themselves without a designer; not only will the designer go out of business, so will all the tools supporting them (big businesses know this, which is why they are forced to get on board quickly before they become redundant).
The way I believe stable diffusion works is you can start with a picture and generate something from it; so you can take a child’s illustration or click sketch and turn it into real art or photograph, basically filling in the pieces.
What I’d love to test is this: cover design is typically, just a 3D render character and a background, with text, some great artists paint over everything and make it one piece, but more amateur designers don’t. I think you could make a simple, quick mockup with a character and background, and then run it through stable diffusion and have it look amazing, any particular style. Or, I think you could find a stock photo portrait you like and type in keywords and styles, like fairy, elf, vampire, and it would transform the picture. (*Midjourney does this now, you can add a photo of yourself and a photo of any fantasy creature, no prompt, and it will blend them).
Also the thing a lot of people aren’t considering, some of this is trying to keep things open source, democratic. Previously companies have owned the tools or output, so you can buy and sell on their platforms or pay recurring fees for their software. For artists, creatives, this is terrifying. But for everybody else, it’s exciting, it’s magical. It’s not fair, it’s cheating… but it’s a little like being pissed off at student loan forgiveness programs just because you paid all of yours back.
Extremely predatory financing has gotten much, much worse, and college has gotten much more expensive. So maybe you built a creative business because you got their first, and school was cheap for you. People who really excel at art probably had a leg up in life, support, connections, opportunities, paying for art school, it’s not all about luck and skill. There’s a certain amount of privilege – I know it doesn’t feel like it at times – but spending hundreds of free hours worrying about the book you’re writing, suffering for your art until you get good enough to charge, that’s a personal choice. It was never supposed to be simple, easy… though it used to be. A few centuries ago, the birth of a novel, there were very few novels being written. Now everybody is writing them. It’s an unfair comparison, but an interesting idea.
The smart thing would be, to create a stock photo site, hire a bunch of people to make “art” and photos, based on what people want/search for. Anybody with a business won’t like this stuff, because it skips over all the access and control issues.
Accessibility and control
People with disabilities, who can finally enjoy drawing or creating, people who can make their own tattoo designs or coloring books. All the things they want to do that they’ve never been able to do before, a whole new magical word is opening up to them. You can’t expect to slow down tools that broaden the means of production and take it away from those in power.
Text to speech is getting so good it will soon be as good as professional audiobook narrators, that’s another complete industry that’s going to fall apart.
Baudelaire called photography the refuge of failed painters with too little talent. Van Gogh said photographs could never capture that human spark, so he started focusing more on painting portraits. The emergence of new technologies influenced our most famous artists, and some were reactive (though, both VanGogh, picasso and almost every great artist in history, we recognize from the last 150+ years… after a thousand years of precision and training and the “right” way to do art, skills taught to few, after Klimt’s vienna secession and the modernist reaction against codified forms of aesthetic perfection, we got dozens of new influential art styles in a handful of decades.
Now everybody without any talent, training or experience can create great art. It’s fantastic for most people; the vast majority of mediocre (aspiring) artists. Current artists, working or not, who have been trying to develop skills, build a following, and find a way to be profitable, this is horrible for them. But art has never been kind to artists. The people who will profit, have always figured out how to make what people want and sell it to them.
Artists have been the inbetween, hired by the business owners, or even something on fiverr, drawing a portrait of their kid or pet. These tools will completely replace the need for them, for most of them. Everybody who can just do the task for an affordable price. It may not replace the established, famous artists, but it will replace everyone else. It’s like, halfway through a race, they call time out and make everyone start over again, and then equalize things completely, removing talent, aesthetics, everything. Fresh clean slate. Those who win will be those who spend enough time building great content and putting it out there.
Every great famous art in history, most, have been anti-gatekeeping.
First poetry, breaking the right kind of meter, using common language instead of stuffy old prose, the iambic meter. Hemingway’s very simple language. Emily Dickinson, the romantic school of everything. Famous poets, Emerson…. Moby dick even. Breaking rules and styles, an act of rebellion. Doing the thing exactly how everybody said you absolutely could not.
The protestant revolution: nailing a list of objections on the church door. One complete oppressive system of the way things had to be done, about what quality is, and needing to pay into that system to be supported by that system, but always to the actual event of fleecing regular people and funneling money towards those in power.
People tend to define art as what’s new; so the only big famous art has been transformative. There are a handful of big famous working artists and then a million struggling starving artists who are not successful, and mostly derivative. Artists almost always start off being derivative, copying and combining, until they find something that’s actually new that people respond to and then they focus more on that to get a unique style, cranking out more work and commissions because they are profitable.
Success leads to LESS originality and creativity, not more of it. Actual creativity, making something new that breaks all the rules, is very rarely a good idea and very rarely profitable, because people don’t like it or don’t get it. In writing, it’s usually a terrible idea, if you’re trying to be an artist. An artist creates for the pure joy and passion (BUT) that only belongs to the amateur, the beginner, who still gets the dopamine rush of trying to do something beyond their capabilities and surprising themselves by succeeding anyway.
Chasing this joy does not lead to better quality work, objectively, though many authors still believe that it does, they are usually the newer writers or the unsuccessful writers. So behind art which is about personal expression, we need another word for everything else; the work. The stuff that pays the bills. We love quotes like this:
‘I Must Create a System or be Enslaved by Another Man’s: The enquiry in England is not whether a man has talents and genius, but whether he is passive and polite as a virtuous ass, and obedient to noblemen’s opinions in arts and science.’
We love to celebrate artists who had it tough. But the truth is, Blake took money from clients and patrons, was generally a jerk, and tried to do his own art on the side (which was a dismal failure). We tend to love the famous artists who died poor even more than those who were successful in their own lifetime.
Imagine, a hypothetical situation, where people can suddenly get something better, for less money. This happens all the time in business. Better means of production. Massive factories and production lines. Something expensive that used to be made custom is now mass produced and artisans suffer.
For example: Amazon deliberately ripping off every successful product on its site, and then selling its own cheaper version. That’s pretty dodgy. But they own the platform.
Of course that’s sad, lamentable, some people will say, angrily, defiantly. Those of means who can afford custom work may prefer to get the “high quality” stuff, even as a mark of status or pride. But what if the new thing actually works better, makes more money, solves the problems. Consumers will also get the best they can afford at the lowest price. True we’ve never seen anything like AI art before, but people are telling me I shouldn’t embrace AI art because….
- Because AI is stealing? It isn’t; not in any new or unique way. Art has always stolen egregiously. Warhol, Picasso, both hacks, blatant stealing. Same with Van Goh, with any artist. They copied, mimicked, studied and chased trends ruthlessly until they found something unique. Warhol is getting sued now because his “art” was just print screenings of real things that already existed, like this Prince photograph. Was it transformative? Not really, but he got away with it for decades and is one of the most famous artists in history. Anyway, if it was stealing, if it was illegal, then the governments and companies would do something about it, to make sure they’re making money without breaking laws… or they’ll change the laws. I have little faith in championing the rights of artists over big businesses. Maybe I’m cynical and jaded, but I’m a realist. My opinions on the ethics of all this stuff, don’t really weigh in on my opinions about what it is or how it’s going to be used.
- Because designers are losing money? That’s sad for them; but industries adapt and evolve, and their cash cow may dry up as alternatives become available. I used to be a passionate starving artist, for many years. So I understand that mentality. But I am not that anymore, I used to beg for help, for support, for friends and family to champion me. I would get frustrated when they didn’t and when I couldn’t sell what I was making. I learned that Art needs to Matter and Have Value; and that’s about market reception, not creativity or inspiration.
- I’m being told I should avoid AI art and not tell people about it because… it’s bad for artists who make a living designing stuff. I get that I sound like a jerk when I say that’s not my fault, I’m not to blame for that. Back to our hypothetical: there is better, cheaper design but I should not use it because…. It’s not “pure”? It’s not “human?” It didn’t take time and effort? Why should I pay more for someone else’s time and effort? If you can make better art in less time, I’ll pay you more for it. But I won’t pay more for less, and I don’t think authors should be expected to do that either. (I’m not really a “capitalist” but I also don’t think artists have the right to demand “support” that isn’t directly tied to value and quality.
You can say there’s a lot of misinformation. I don’t think you can say I’m uninformed or not paying attention. I don’t think you can say you understand the copyright law better than I do. Which means, you object on a moral level, and you’re directly impugning my character. You’re allowed to have your own moral ethics, you’re entitled to your feelings and opinions. But when you make blatant statements like “how can anybody support this!” – you’re saying I’m a bad person; that I’m doing something immoral or harmful. I don’t feel we can have any kind of discussion or conversation if you’ve got an implicit moral bias against me; the facts don’t matter in this case at all, you’re determined that this is bad. Now, is it bad for everybody, everywhere? Or just for you and other artists feeling threatened?
Some have said, nobody will profit from this but the big companies. Yes! Big companies are bad, for sure. But they make money because they provide things people want or need. If we go that route… why are we using any software, any computers, anything at all… money always flows up to the top. What do we gain by refusing to use new tools that everyone else is using?
A friend pointed out that artists are (rightfully) upset about this, so for that reason alone we should all stop using AI art to reduce their anxiety, and I really appreciate that idea. But how? Someone in the comments called for a total 2023 book ban and saying “Hope everything you’ve ever done ends up on pirate sites and scraped for millions of other peoples AI garbage” (this was one of the kinder messages I’ve received).
Lighting Round (FAQ)
Will people use them? For sure.
Will companies keep selling them? Also yes.
Can protest sometimes effect meaningful change? Sure! I get it, protesting is all you can do. Someone should be protesting and holding companies accountable. But if it’s not illegal all you can do is color public sentiment, which again, that’s your perogative.
IP theft is bad. Therefore AI art is bad.
But this read is a failure to understand the technology or copyright laws.
Is it a threat to artists?
Is IP theft bad?
Will this screw over everyone except the big businesses?
Yes... except the millions of people who happily, excitedly use it.
Is it legal to look at copyright protected images and be inspired by them to create something new?
Is it illegal for me to use these new tools?
Is it currently illegal by any stretch under current copyright laws?
Should it be illegal?
Sure, probably. From a philosophical, mankind is doomed point of view. But if you don’t think mankind is already doomed, by much bigger problems, you haven’t been paying attention.
Do I think it will be made illegal later?
I do not. This is not a moral judgment. Maybe it’s a lack of optimism. Maybe I should fight for artist’s rights even when it’s a losing battle. Maybe by enjoying myself with AI tools instead of sweating it out through labor and equity or always paying for whatever I don’t have the skills to make myself is the only right (moral) way to handle this and I’m revealing myself as a selfish lizard creep, not a real artist. I can entertain that point of view, certainly. I’m also fairly certain I can freehand draw or sketch better than most, after training in classical fine art, though my comparative skill level is meaningless to the discussion.
Neither my level of skill, how long I spent honing it, how much I spent on art school, nor how much of my income is currently threated by AI tools, matters at all in the general discourse, at least to me: though apparently these are the reasons that matter to other people who are angry about it.
Is everything legal good?
Of course not. Alcohol is bad. Smoking is bad. People like to use them anyway. At one time the government decided to let adults make their own decisions (except during prohibition). But not for seatbelts (we have to wear those). Laws are weird. And no of course this example isn’t a good fit for the AI debate, I won’t pretend it is.
But it won’t matter what arguments I make because arguments don’t matter.
Is moral grandstanding and righteous indignation the *only* way to combat the threat, which isn’t illegal but maybe should be? Sure! It might even be the smart business move to rally troops around you. I’m not smart at business. I’m practical. Easy, cheap, fun new tech that augments humanity’s creative abilities to produce better content with less effort, skill or training, is something I can be excited about.
I’m a factually wrong with the idea that it isn’t currently illegal?
No. Am I morally bankrupt for using it anyway despite concerns?
Depends on which ethics we are talking about.
I gave up the starving artist mentality years ago: that real art is precious and passionate; that real art has value because of how hard the artist struggled with it; that real art has a soul that has value even if nobody wants to pay for it. That was a tough, brutal lesson after years of being a struggling artist turned out to be not so much fun, and I realized I was just begging people to support me without offering anything concrete in return.
I’m a philosophy major; I’m pretty familiar with the history of ethics, from Aristotle to Bentham.
Most boil down to “do no harm” and the “greater good.”
Will this tech harm artists? Yes.
Are there *more* people whose lives will be improved or enhanced? Yes.
Do the opportunities outweigh the harmful effects? Maybe, maybe not.
Do ethical questions like this *ever* stop or slow down the practical adoption of new techologies?
Is the world all coming to an end because humanity’s core ability to create is being replaced by robots?
Sure; though I think the world is ending for greed, corporate profit, the absolute destruction of any kind of shared reality; the absolute pollution, destruction of ecology and nature, consumer culture, social media, a million other things that I personally have no control over.
It’s like recycling. I know I should do it. I believe in it. I support it.
But I also know recycling doesn’t actually work; that it’s a lie constructed by companies to make me feel a little better about using disposable convenient packaging even though it’s harmful; they are mitigating my feelings of guilt. But I know the lie, so it doesn’t work anymore. I feel bad out it, hopeless, but I know my little bit of effort has no universal effects on a greater scale.
Am I automatically in favor of IP theft and anti-artist because I acknowledge these tools exist, are fun, easy and cheap, produce great images and drastically speed up my production time while giving me better (more competitive) images that look great?
Most things that actually make money are not art, nor are they creative. Just because a human made it does not make it more beautiful, more valuable. No matter how much sweat or passion went it to it.
Midjourney makes beautiful, emotional art – it’s funny to say it has no soul, but it still generates an emotional response when you look at it. Does it matter if the robot doesn’t understand or feel; or just that the receiver/audience does?
How about sites like Squarespace and Wix…. They are not “real” websites; they’re too easy, not enough control, just pick a template. Looks great amazing/better, less time and effort, less money. But they’re “stealing” from real websites made by web developers/coders and it’s not fair, nobody profits but these big businesses who are selling easiness. So you must be lazy/unethical/cheap to use them when they are stealing work from real developers who spent years learning web design.
To wrap up, here’s some email comments I’ve gotten from people who are enjoying AI art.
“It is probably the best possible thing that could have come along to streamline my book production pipeline.”
Your email changed my life. I wasn’t going to send you this comparison picture because I was mortified when I re-visited a cover I made for a friend over a year ago, but I know you must get way too many emails, and nothing can really describe how much of an impact AI art has already made for someone like me than this side by side example. I’m not an artist, and I’m not even a good designer either. But I really enjoy making covers and concept art, even if that particular cover had me wanting to throw things at walls when I was trying to put it together. I’d like to think I’ve gotten a little better since then.
I spent a day scouring your email for every link and was already generating images as I listened to you talking. I just wanted to say thank you. The free content you put out all the time is incredibly generous, and I wouldn’t be where I am today without having learned back in 2020 how to make a cover in Microsoft Word with one of your tutorials.
“I really look forward to seeing more of your AI art—what I’ve seen so far is out of this world.”
First of all, I’d like to thank you A LOT for all the amazing resources that you create. I just discovered AI art with your last email and I’m so thrilled by it! I made a few attempts with Midjourney.
A deeper cut (more comments)
These were taken from another post, but good points were raised…
How the image based current systems play out will set the precedent for the fields that follow. Which is one of many reasons they should be boycotted.
Interesting theory, I totally agree with you and this is a smart argument. I’m just not hopeful enough that we can/will forbid AI tools from selective personal fields of creativity (and then judging which fields are creative or noncreative; which jobs can and should be replaced). That feels messy. But it’s true we should pause and make sure somebody is paying attention, which is what artists are doing now, and I applaud them for it. But boycotting, that’s a complicated word. There are many companies who are on board. Who are we boycotting; the default AI software is already open source. Apple, Adobe, Google, Amazon? You’re going to boycott them, because they’re already invested. It’ll become part of your everyday experience.
The value of a given service or skill is proportional to its scarcity. Once your hard won ability to write and market a novel is no longer scarce, its value will plummet. And it will be gone, along with the fulfilling life you have built around it.
Ouch. But your premise is flawed (generally, though specific with book cover design). Here’s the thing: AI will get good at writing and publishing novels. As someone with a PhD whose spent the last 5 years writing fiction, I would be gutted if robots suddenly did it better. But they won’t. It’ll be some kid using AI ghostwriters and AI art to crank out passably good novels. Which would be terrifying… if I didn’t already plan to stay competitive and do that myself. All artists can do at this point is become familiar with the new tools (which happens for everyone in every industry; though this specific thing is more powerful and disruptive than anything else) or refuse. Nobody knows what the future will bring but it is here already, only some of us are paying attention.
The reason art is treated differently is because this is what we do in our lives for meaning. We weren’t supposed to replace all the fulfilling jobs and leave only the drudgery, that’s the wrong way around. The end game is this comes for everything, including you. Do you honestly believe we should have no musicians, no actors, no authors, no artists? Is that a world you want to live in?
This is a romantic ideology that has saturated the modern world (that art has more value especially when it has no value).
Art is more valuable to humans because we can feel it. YES.
Art stops being valuable in this way (to creatives) when robots can do it faster. MAYBE.
Part of the joy in art is improvement and becoming the best, and robots threaten that. MAYBE.
AI will replace all creative endeavors and do it better. YES. (Art hasn’t produced much new in the last few decades, I would argue, precisely because too many people are doing it and we are all too connected).
AI will allow creative people to do bigger and better things than we’ve ever been able to do before (YES).
Artists will refuse to participate and opt out of society like in Atlas Shrugged: MAYBE but they won’t stop the heart of the world, because the robots can take over.
“Do I believe/world I want to live in.” Honestly I’m reporting what’s happening in the world and introducing new tools. I’m like Brad Pitt when asked about Tibet – “yo I’m just an actor dude.”
I understand by not taking a negative stance I’ve become controversial in a hot topic that artists are eager to shut down, but the tiny part I play in the social discourse of creativity is pretty small comparatively, and you could go after much bigger fish. Am I a target because I’m an author/artist and I should know better? I get the reasoning. I’m uncomfortable under the rage and ire, and fearful of mob mentality, which I regard with respect. I’m not against artists and I’ve been pretty polite in my discussions. Yet here we are, I’ve become a strawman. Light me on fire.
You should be boycotting this, unless you will be equally keen for the writer replacement bot to destroy our industry.
Yes, I should be. Like all other content creators for creative people. It’s the correct moral stance right now. But boycotting emergent technologies has never gone well; especially ones that increase access. I’m happy for all the non-artists who are super excited about being able to do things they couldn’t before. I’m sad for the artists who will need to recreate their income strategies. But I’m not angry… angry at who, the robots, the big corporations? Since when have they listened? I can already buy fully stock AI images on big stock sites and use Dalle2 inside of photoshop (I think, or something like it… photoshop has a bunch of built in AI tools that make things WAY easier for designers now than when I started 15+years ago).
Most creators LOVE ai tools, like stripping out a background or replacing the sky. They make production easier and faster. That’s great for them, up to a point, and the point is whatever service they are personally charging.
Am I keen to be replaced? Not especially. Do I think my opinion matters to the billion dollar companies that we pay to do everything? Not at all.
There is no winning argument here. I’m either for the bots, or corporations, or anti-artist. Even if I’m not taking a stand, that’s taking a stand because I *should* be using my platform to raise concerns about this incredibly huge issue. Sure I get that.
But I’m not going to. Partly, because like the meme I shared above, a lot of it is pure disinformation. I can’t ignore it, and I’ve never been good about condemning things (my graduate studies have all focused on Milton, freedom of speech and press, and technological enhancements of creative fields). I love AI because it’s amazing, full stop… I understand the controversy around it. I’m not saying artists shouldn’t be threatened of freaking out about it.
Percy Bysshe Shelley wrote a passionate call to arms in defense of Greece (Hellas, 1821) which motivated Lord Byron to go and fight to protect the Ideals of Freedom “(to produce fresh generations to accomplish that destiny which tyrants foresee and dread”). 3 years later, April 1824, Lord Byron died in a fit of fever in Messolonghi, where he had gone to aid the Greeks in their fight for independence from the Turks (probably infected from unsterilized equipment during a malaria bleeding).
We champion his brave gesture and the motivations behind it. Byron became a hero, and after 11 years of war, the Greeks won. But time moves so much faster now. By the time you’re starting to hear about these AI tools, the outcome has been decided, the war is over. I’m not asking you to lay down your arms. Just try not to kill anybody.
AI writing predictions:
I’ve had a lot of people feeling worried about the future in general, and also people who were initially curious and now afraid.
– I wouldn’t be surprised if Amazon starts writing their own books based on the popular ones, like it’s already done with popular products rebranded for their own line.
– Amazon will probably try to ban pure AI written content (or not, they might jump on board especially if they are using/selling it also).
– Writing insanely fast may not be the hack it used to be… I’m a slow writer at 4 books a year, so I couldn’t compete with people cranking out books. This might help.
– The words don’t really matter, they are just filler for the story. AI might help you get more better words, but the story is what will be remembered (you can have AI plot/tell a story, even map out a full outline though).
– I think authors are going to be able to do much more themselves, without having a team of editors, designers, betareaders, etc – and be able to put out better books, faster. But the tools only help if you know what they’re doing and give them a purpose. Good writing isn’t what makes a book good, so you need some high level of story craft and they’ll still need a lot of revision and planning.
– AI will never just write books because they need a human agent who wants to publish and promote and make money with them. Most new authors will play with the tech and find it fun but not really know how to build it into their own writing process; some people will try cranking out books but they won’t be good and they won’t know where to start with cover design, marketing, etc.
– But some people, who already have experience, might use the tools to enhance their productivity (quillbot for example, rephrases your writing into different styles, so it’s like grammarly or prowriting aid but actually makes changes instead of flagging a thousand things.)
– Right now it’s too heated to really do much of anything, but soon everybody will have access to tools that never existed before. It’s probably the end of the world, sure. Do we play violin on a sinking ship or build a lifeboat?
– People are upset about AI art and swear not to use it, but Adobe and shutterstock already accepted it into their system… I really get wanting to support artists, and not using tools that harm them or make them upset… but where are we going to get our stock photos from? Why use grammarly if quillbot (a more powerful AI tool) or just chatgpt3 can edit our books better, even improve the style and tone?
The options are: use it or don’t, whichever brings you peace, whichever helps you get your best work. Arguments about why nobody should ever use it and it should be banned… I feel like the train has left the station. Who are we going to ban, the internet? Should we go back to writing on typewriters? Write by hand?
Anyway, I’ve written a ton on this stuff and I don’t think the heated debates are useful, and I’m emotional fragile because of all the hate I’ve gotten around AI topics, so I’ll mostly be staying quiet and playing with the tech.
But I’ll leave you with this comment from author Jamie Butler, who is using an AI feature inside of Notion. I think in a year, a lot of/most writing apps and software, including Word, will have some version of this: basically like “clippy” – a friendly assistant who can do everything and answer questions so you don’t have to go searching.
“It seems that the application for writers is endless. Aside from the obvious utility for idea generation, drafting, editing and summarization, I’ve found the AI super useful for research, using it as a query tool. Even prompts such as “Suggest 10 ways to curse in fiction without actually cursing” and “Suggest some similar words for ‘scintillating’” have yielded helpful results. It’s like having a knowledgeable and helpful “Jeeves” hovering at your elbow at all times, waiting to answer whatever you ask it. I have only just scratched the surface with discovering what it can do, and I’m excited that it is there, integrated into my own writing workspace. This does bring up the topic of what AI could mean for the future of writing (and writers). I’m very curious to follow the conversation, and I hope you keep talking about this in the future!”
I’m not pro-AI. I’m pro-artist, and creativity, which includes innovation and new technologies. I’m pro-independence and individuality and personal expression. I’m against gate keeping and excluding the right to the joy of art and aesthetic discovery, to people without access to the time, training or resources to become as good as they want to be to achieve their creative vision. I’m opposed to pulling up the ladder behind me because I’m afraid the riffraff will get in using cheats and hacks, instead of honest hard work (creative people have always found ways to get better results, even going back to early uses of a simple projection machine that much later became the camera, to create near perfect paintings.)
Mostly, I’m a kid in a sandbox, and a tsunami is coming. I’m just looking up and asking how it will affect the castles I’ve built, and what I’ll do when they are washed away.
I’m a philosophy dropout with a PhD in Literature. I covet a cabin full of cats, where I can write fantasy novels to pay for my cake addiction. Sometimes I live in castles.
You raise many important points here and it is good to hear a more realistic and differentiated view. Most people at the moment are far too emotional to look at the facts and to defend their views with arguments other than “AI bad, humans good”. I am an artist myself and I would count myself among the mediocre, and it hurts to say but I think the work I have been paid so far for my “skill” I have always found overpaid anyway. It was only a matter of time that something like this would be replaced.
this is stuff that spark a war
This was a fascinating read! I am witnessing a technological revolution in art and creativity. It is so interesting to me from a psychological perspective to see artists wanting to restrict creativity. I am truly captivated by this topic and am excited to see where this all goes and how much better(or worse depending on your view) it gets in the coming months/years/decades. When artists have to talk about what it means to create “art” you know that this technology is beyond anything people could anticipate. I do believe the cat is out of the bag on this one and there is no going back.
Super insightful read. I’m mostly in agreeance with you on everything, the good and bad though I can’t seem to quell my anger despite being an awful armature artist. The only thing I think you might have over looked is the glutton of garbage that is going to inevitably flood the markets like never before. It’ll be Youtube all over again with terrible vlogs and reaction videos.
I loved this article, an as an artist myself I am having a hard time, being bullied and loosing friends because I gave permission to myself to do some experimentation with AI. This sounds crazy to me, and not good for the cause , too.
(I am a foreign reader, sorry for the language)
Let’s also say “the war against guns has been lost, so don’t try to prevent school shootings” or “the war against drugs has been lost, let’s not do anything about overdoses.” Why should anyone fight against the evils of the world? It’s easier to just go along with them, right?
Spoken like someone who’s never had to fight just to stay alive.
Sure, that’s not a bad argument. It makes sense to fight against powerful forces, even if the battle seems lost. Mostly, I see one side using logic and fact, and one side using emotion and disinformation. The argument against AI art doesn’t seem to have any solution. I don’t believe an outright ban will happen, and there’s no legal precedent for forcing any changes. As an artist it feels overwhelming and scary. Everyone is free to respond how they want.
Yeah I find myself leaning more towards Derek’s sentiment. To use your analogy – even if one discards the fray that may feel school shootings are an outlier occurrence, greatly overrepresented in our public psyche and not a priority in the grand scheme – and focus on ways to diminish their incidence – different people have different views on how to prevent them. Some will focus more on culture, on safety, on respect, and on underlying mental health. Others will focus on power dynamics, on supply-lines, on regulation, on safeguards, and punitive measures. If you’re a person that doesn’t fall securely in one camp or another, it’s very difficult to ‘fight’ your perceived evils.
Ironically, it’s supposedly questions like these that AI is being championed by some to take-on and provide solution-sets for… But since the core of these aims will derivate from existing [and selective] stock of persuasions, it’s not wholly clear that this will be useful; it may very well be counter-productive, as a layer of ‘outside authority’ (AI) will be imposed on us. It’s like citing an incorrect sale price with the cashier, who says, ‘sorry that’s what the register says’ contesting a ticket with the traffic cop, and they say, ‘well it’s already in the system’. You can fight it, but each battle is of negligible success, if that. Have you ever tried arguing with a store that won’t take cash, to take your money?
I’m still in the camp of my opt-out choices being the most tangible way of casting a ‘vote’ (when you think about all the sentiments of ‘democracy’ we supposedly champion, it’s telling how paradigm shifts as profound as AI or genetic modification are not directly provided a say; reliance is shifted to politicians to be mediums in these debates)… But I’m a dinosaur that will likely be relegated to near-extinction, or need to magically morph into a dragon, or rodent, or something more fitting..
Regarding various peoples, who may or may not have to fight for their livelihood and culture to prosper (or change), I found this article chart interesting. Such vast discrepancies – but clearly the US (who is developing a good chunk of these technologies) is the least trusting of them. (I’m not sure 20K people across the entire globe is a worthy case study, but I’d guess it gets at something):
This was well written and provided a lot to think on. There’s a lot of philosophical (and legal) questions related to AI art, and while the legal questions will eventually reach an answer, the philosophical ones likely never will.
Personally, I believe in the artist and humanity. I very much hope and believe that we will continue to create, even if AI art and writing becomes indistinguishable from human art.
I’m not going to express all of my opinions in this comment, but I would like to thank you for this article. While I disagree with some of what you said (not most, just some), you provided a lot of insight.
Thanks I appreciate you! It’s messy and complex. I could be wrong on most things (I agree it’s terrifying for artists and authors, and bad for humanity in general… unless it isn’t – we’re pretty much destroying ourselves as is, maybe AI will provide solutions we couldn’t come up with). For creative stuff though, it’s both a huge gift and a curse. So much easier to finish my best work without all the pain and struggle; but that also means it’s easier for everyone else, even people with no skills or training, who didn’t spend a decade improving their craft. I have mixed feelings to be honest, but it’s kind of too amazing to not be using, and I think most creatives will find a way to improve their workflow, creativity and productivity.
This is not targeted to anyone in specific.
I keep hearing about disabled people…. As a disabled artist I don’t find it particularly amusing. If someone can’t draw because they say, have a condition were their hand shakes too much, the situation is the make a medical device that steadies them. The talk about ai making art accessible to disabled people uses ai as a fix-all for any kind of disability imaginable, while ignoring disabled artists who already draw like myself, and disabilities that don’t profit their argument as much like, say, being blind.
On top of this, their attempts to “democratize” ai for disabled people are really funny, considering most disabled people are extremely impoverished due to difficulty getting and keeping jobs, and the high cost of medical bills. When something like nijijourney costs 30$ a month….. How many disabled people are being “democratized “? I really wish people would stop pretending to care about disabled people.
Pro tip, if you want to help disabled people, don’t use them as voiceless agentless excuses for committing crimes of theft. Instead, consider donating to disabled organizations and charities, volunteering in the community, and doing outreach.
I’m glad you take into consideration the perspectives of those of us who aren’t artists. The power to bring your vision into the world is amazing, but for those that lack the skill, a middleman has always been required. I give a text description (and sometimes a reference image) for the vision in my head, wait however long, then give more text to try to get the image to look like what I’m imagining. That repeats until I’m satisfied or run out of money, essentially. Whether the middleman is a human, a corporation, an AI, or even an alien doesn’t really matter to me. What matters to me is the quality of the end result, and how much time and money it took to get there.
ah… and this is the part — however pervasive (or, perhaps because its pervasive) — that brings the sadness: “Whether the middleman is a human, a corporation, an AI, or even an alien doesn’t really matter to me. What matters to me is the quality of the end result, and how much time and money it took to get there”.
From many standpoints, this is where the doom has lied for some time. It presents, literally, the most alone and selfish conception of individualism that one might take.
At that point, I [truly, not snarkily] wonder: why does the quality even matter? Do you care if the quality is being consumed by a human or corporation? Do you care if your family is an alien or an AI? Where and how does one begin to ‘care’ about something? If its all about money, why would AI even give you any, for your ‘quality’ X product?
By the way: well-written piece, DM. A cut above.