Introduction: Cute Anime Girls and AI Content Creation
If you’re a writer or an artist, you’ve probably seen advertisements for Artificial Intelligence (AI) tools that’ll write blog posts or create graphics for you. I’ve seen lots of them. They’re apparently popular. But honestly? I can’t understand why I would want to use them. Until now, I really couldn’t articulate the reason. Well, I think I’ve figured it out, and I owe it all to cute anime girls. And not in the way you might think!
Or maybe exactly in the way you might think.
Important note: This post is for adults – as in, eighteen years and older. If you’re younger, please do skip this post. The content is not explicit, but it’s intended for mature audiences.
AI Images: Sure, They’re Cute, But…
My inability to speak my mind started with a proliferation of AI-generated anime-style graphics on Instagram. If you’d like an example, check out the account called uki_hajime. Now, before I say what I want to say, I want to make something clear: I don’t dislike this work. I don’t want to disparage it or insult anyone who loves it. This is just my take as someone who creates content – and who just figured out why AI content bothers him!
Take a look at these randomly-selected images.
Are these girls sisters? If not, why do they look so similar? From left to right, here’re the links to the original posts: Light armor, dark armor, and light dress.
I kid you not: when I first started seeing these, I thought they were supposed to be sisters. They all had similar eyes. They all had slender shoulders. And…
Well, before I say the “and,” I need to set the stage for an idea that turned out to be critical. It has to do with plot and backstory.
Yes, that kind of plot and backstory.
Pixiv: A Source of Human Inspiration (and Probably Some AI Content Creation)
If you look at an anime art site like Pixiv, there’re probably a few things you’ll notice right away. First, the art there looks like authentic anime art. Most of the work could drop into, or be lifted out of, popular anime titles. I love that style, and I love that site. In fact, when I write my novels, I’ll page through the day’s popular images and find one that speaks to me. She (yeah, I’m partial to cute anime girls) then becomes my muse for the day. Here’s a portion of one I used recently:
You can see the whole original image here. The artist, Dino, has a beautiful, wistful style that appeals to me.
Now, that’s a perfectly suitable-for-work image. There’s something about it that I couldn’t put my finger on – some idea that made it appeal to me more than the AI-generated images (at least, the ones I can tell are AI-generated). During one of my recent muse excursions, it dawned on me. It’s something else that will hit you if you browse Pixiv, or if you watch anime in general.
Have you noticed the plot in some of these shows? And by “plot,” I mean breasts. Some of these women seem to have breasts that are larger than their heads. As in, by the time they hit thirty-five years of age, they will likely need back surgery. That’s not to mention the struggle they must go through finding clothes that fit. But in general, breasts figure prominently in anime — even the “flat is justice” movement.
Other shows or artwork focuses on backstory – which is to say, shapely behinds. I should probably say “asses,” but I’m trying to be higher class here. But the way the image focuses on those curves makes it clear – this artist thinks that aspect of a woman’s body is very, very (did I mention very?) appealing. There’s nothing wrong with that – just as there’s nothing wrong with focusing on breasts that are big, small, saggy, perky – whatever. It’s a human observation communicating a human reaction.
That’s the key.
What This Art Has to Say about AI Content Creation (Indirectly)
Look, I like plot as much as the next guy (or girl). But, dang! Here’s a link to the original image. The artist, 朝凪＠新刊出ました, maintains a page here.
What are these works of art telling us? I can’t speak for any given artist, certainly, but I suspect those artists intent to use their art to communicate. The communication might be as simple as “breasts are beautiful,” or “hips contain the meaning of life.” The art reflects the writer’s interests — or the writer’s kinks.
Back to the AI art. I thought those girls were supposed to be sisters. One of the reasons was that they all had realistically sized breasts. No deployed airbags! There was no hint of what the “artist” liked about these girls. They were competently created images, using competent models, with competent composition and lighting. But they weren’t the product of any human kinks – or any human affections.
That’s key. Because I browse Pixiv looking for cute anime girl muses, I experienced a revelation that was really right in front of me all this time. Artists communicate through their kinks. I could then generalize that idea to become: one human communicates a human idea or sentiment to other humans.
When I look at the human-generated line drawing above, or the drawing of Echidna, I sense a human connection. When I write novels, I try to include something of myself in those works. I try to initiative that connection. I try to communicate ideas, or get points across. Visual artists are the same way.
That’s what I’m missing when I look at AI-generated content. Yes, it’s beautiful. It’s even technically correct. But there’s no human connection. The lights are on, but there’s no one home.
The Assumption in the Room
Now, I’m making an obvious assumption here. Namely, I’m assuming that the artist has something to communicate. It may be that a blogger simply wants to drive traffic to ads. It might be that a graphics artist wants to drive clicks to his or her page. I’m not disparaging that. People gotta eat. But that’s not what I think art should be. Nor is it what I think artists should do.
It’s not just limited to plot and backstory. There are other more complex ideas – and some even encourage a wide diversity of figures – and species affiliations! Like this picture. You can see the original here. You can see the page for the artist, SAMIP, here.
And you know what? That’s what troubles me about AI generating content. It’s not the problem; it’s a symptom. When I read a post, or look at an image, I want to understand a different human’s perspective. I want to see the world as they see it. That might sound strange given that I started off talking about cute anime girls. But that was just the gateway thought for me. It helped me realize why I had reacted the way that I did to AI-generated art.
Am I saying that we shouldn’t use AI? No. In fact, I think AI can be a fantastic helper! Having an AI find grammatical mistakes in my novels would be fantastic. Having an AI point out that I changed a character’s name by accident halfway through would be great. But in that case, the AI is helping me communicate. The AI can help remove impediments to my communications. As an assistant, AIs can be useful.
Human Content Creators Create Content for Humans to Consume
But as content creators? As long as I want art to communicate human ideas from one human to another, then I don’t think AIs should have a central role. At the very least, I have no intention of ever using one to write content for me. The only thing I have to offer you here on Crow’s World of Anime is my distinct perspective. You can’t get that from any other anime site (or any other novels). If I have an AI write for me, what’s left? Why would you visit my site and not another?
What do you think? Do you enjoy AI-generated content? What role do you think AI can play in the context of human creativity? I’d love to hear your thoughts in the comments!
Note: The featured graphic is from Pixabay. You can see the artist’s profile here.
That’s an awesome picture of Echidona!
There’s actually a deeper issue, I think. Digital technologies have made life easier, and it didn’t start with AI. I’ve seen CGI take over first video games, then animation. I remember what a big deal in CGI the movie Final Fantasy – the Spirits Within was. Nowadays we expect video games to look like that, if not better (I must admit I don’t have the eye for it). And right now there are human beings who’ve come of age and were born after that film came out. It changes how you look at things.
I’m not confident I could tell an AI generated image from a CGI rendered one. I remember watching the anime Earth Girl Arjuna, and they had a hand drawn ending for one of the last episodes (for all episodes? It’s so long ago…). And I remember wishing the show would look like that (but it’d have been far, far harder to animate like that; you tend to simplify concept art for animation). I love the hand-drawn look; I love the irregularities in there I think.
Digital tech tends to remove the irregularities from the process. Everything’s glossy, shiny. Take things like autotune, for example. You can snap things to pitch and beat, fully regularly. No person can sing that perfectly. Should they? I often wonder what it’s like if you grow up with pitch-corrected recordings on the radio; that would be when you learn to appreciate songs. How would you deal with irregularities? Where I intuitively hear expression others might intuitively hear mistakes?
People have used AI to make covers for books or records. Would I, were I self-publishing? It’s cheaper, and given what a paper-back in 80ies might have made from my book, it wouldn’t be worse at least. I can refresh the button and tinker with the prompts until I have something suitable.
This is a strange new world for me; but plenty of people have grown up in it. I sometimes CGI animation looks like to people who were born the year Shrek was released. I mean I’ve seen computer graphics evolve from Pac Man and Space Invaders to… whatever it is we have now (I’m out of the loop, as I literally couldn’t see the improvements after, say, The Last of Us). People born in the early 2000s haven’t. It’s worth thinking about.
That picture of Echidna made me write a couple of novels! I really like that artist’s style!
You bring up a very good point regarding being able to tell if a given image is AI-generated. The examples I chose for this post were self-identified as AI-generated, so it made things easy. I agree with you — in the wild, there are times I’m consciously not sure. There are other times when I’m pretty sure I should be be doubtful but don’t know to be.
That casts into confusion our human inclination to interact with art as a means of communication. It’s like an abstraction on top of an abstraction. I haven’t traced the implications yet.
I’ve been thinking about your points regarding digital tech making everything glossy and autotune’s impact on the vocal arts. I read a short story once where a mediocre musician underwent what was, at the beginning of the story, a cutting-edge operation. He obtain pitch-perfect hearing. He started to move up based on that alone.
Soon, though, everyone had that same ability, and he was back to the middle of the pack or lower.
It’s not a race, is it? Art isn’t inherently capitalistic, is it? Maybe I’m engaged in wishful thinking here. Maybe my hope to tie art to human communication is itself an anarchism — or maybe I’ve misinterpreted history, and it never really existed.
When it comes to video games, I can accept CGI as art because I know what goes into the creation of wire models, textures, and other digital artifacts — all under the direct control of humans. Would I still enjoy it if an AI generated it?
Would I even know?
On the other hand — I’m working on the second draft of Evolution’s Hand Book 4: Blind Exodus, and I’m going through the chapters with ProWritingAid. It uses AI to tell me where I screwed up. I have to say that I’m making that poor AI work its butt off. When it makes suggestions for changes, the first thing I ask is if it’s technically correct. The second is if it makes sense, given my voice. You mentioned imperfections — sometimes, I leave an imperfection in place, on purpose, because it better represents my voice.
After all, Tolkien’s work is jam-packed with dangling participles.
A couple of notes:
Re. Short Story: I haven’t read that short story, but… I’m dubious of the concept. Perfect pitch gives you fairly little advantage (you could sing from sheet music alone, for example, without accompaniment; that would be one). I doubt suddenly acquiring perfect pitch would make more of a success. It’s just one skill among many, and not the most important one. And even if you’ve got all the skills necessary, you’re still not guaranteed success.
The problem with ptich correction software is that, if used indiscriminately, it removes expression parts from the voice (vibrato, for example). The idea that more regular is better is just… dubious beyond a certain point.
” When it makes suggestions for changes, the first thing I ask is if it’s technically correct. The second is if it makes sense, given my voice. ”
That’s definitely the best way to go. You *need* your own voice.
“When it comes to video games, I can accept CGI as art because I know what goes into the creation of wire models, textures, and other digital artifacts — all under the direct control of humans. Would I still enjoy it if an AI generated it? ”
My sister has studied archeology. She once showed me a picture of an excavated stone with holes in it that looked like a smiling face. The archeologists couldn’t determine if the wholes were made with a tool, or if that’s just how it looked when it got picked up. Question: Does it matter?
I definitely accept CGI art as art. What else would it be? Whether I like it or not has little impact on it.
“That’s definitely the best way to go. You *need* your own voice.”
You know what’s weird? I’ve been writing since I was about 10 — so, for about 50 years. By this point, I should be able to listen to myself when I say, “The software’s suggestion isn’t right in this case.” But it’s still an effort of will.
“Question: Does it matter?”
That’s a good question. I think the answer is, “it depends.” I recognize that some folks produce artifacts — novels, graphics, music — simply to make money. Some try to create something for the sake of art. Others want to eat, but want to communicate an idea, or a theme. In the case of the latter two, I think it does matter.
But I think I’ve just shifted the context of the question to the side of content creation. If I just look at it from the perspective of a consumer, does it matter? If I want to just spend a mindless 30 minutes on an episode, or look at a pretty picture, of have an audio book going in the background, no, it doesn’t matter. If I want to engage with philosophical ideas presented within a narrative framework, or within the considerations that go into a graphic image — then I think it does matter.
Unless the AI becomes sentient, in which case, it might an interesting exchange.
The thing that really bugs me about AI generated anything is it seems like people are using it not to enhance creativity, but to replace it. Artists, writers and musicians already work long hours and struggle to make ends meet. We do this because we’re passionate about our craft and we want to express our deepest emotions, to connect with others. Imagine working for hours, days, weeks on end, pouring your heart and soul into a project. And then you see some AI generated thing that took five minutes go viral. It feels insulting, like everything you did was for nothing. I don’t mind if people use AI to brainstorm or do research for projects, but why in the world are we outsourcing creativity — one of the things that defines us as a species — to an unfeeling machine?
That’s a very good point. I didn’t look the situation from that angle, but since you brought it up, I’ll say that might be my biggest concern.
Tech should liberate people. I’m really intrigued by DerekL’s democratizing insight, and I think your comment enhances it. Using AIs to replace artists is anti-democratizing. It’s concentrating power in the hands of those who are first to exploit it, or already have the widest reach. It’s not helping artists sell their work.
I remember when the web got its start. I remember feeling excited by the possibility of people sharing information freely. And of people correcting errors freely. We see how that turned out! I’m afraid AI is heading in the same direction.
As both an artist and a writer and someone who is close to a number of musicians and song writers as well – this is my main concern. It is pretty amazing what AI can do. But the capitalists who run the industries creative people work for (or hope to work for, or at least freelance for now and then) have long bemoaned the fact that they have to pay for artwork, or stories, or music, wrriten by those damn creative people. They pay as little as they can and have entire departments devoted to then screwing the creatives out of what they did get in pay. We live in a society that devalues creative work and creative workers as “lazy” or “do nothing” because they don’t see that drawing or writing is hard work. To me, I feel like these capitalists are jumping around in glee that they can now get rid of an entire group of people they hate to pay and replace them with an AI or two. I’m out of the work force now, and glad of it.
That said – two things I think. One – I played with an AI art generator for free for a bit on a website, and it made some pretty things. Some I even liked. But many, most even, never quite came up to my mental visual picture. What I see in my mind and what the AI put out were always not quite – and to me, and my vision, my vision was …well…my vision. And I think on that basis, I agree with your point that AI generated work doesn’t have that human connection. That lovingly curved ass.
When I was writing sex stories for pay there was also at that time what was called a “story spinner” being used by some pron site web builders. It would take a bunch of sex key words and swizzle them in it’s software blender with connecting words and spit out a “story”. By and large they were beyond laughable and made no sense at all, but they’d work as click bait and maybe the clicker would click out on one of your sponsor ads and get you a few pennies. But believe me, you could tell there was no human element in those stories. I just read an article saying that Amazon was recently having an issue because someone cranked out a few million young adult romance “novels” with the AI equivalent of a story spinner, and again, they barely even make sense. But as a writer, even sex stories, I write with making some sort of point in mind – even if I don’t have it in mind. My personal ethos about sex is going to permeate the story and this is true of any author. Again, I agree with you that no AI is going to write with a conscious or subconscious agenda. One of my favorite authors is Dean Koontz. Not for his novels, exactly. I am not a big fan of horror. But because he has a way of inserting into the scariest story profound observations and thoughts on life. I just don’t think we are going to see and get that from an AI. Oh, and Tolkein does the same in a most whimsical way.
So I agree, I don’t think AI is going to be able to CREATE the way a human does. Ever. Unless they do someday become sentient. But I do think we will see them doing more and more because the profit mongers will jump with both feet in great joy on the chance to get rid of paying creative people. Consumers… I think that is an individual thing. Some will notice. Some will not. Some will mind. Some will only mind some of the time. Think of the big CGI controversy in anime. Me? Meh, I don’t mind it. But I also appreciate the way older anime use other methods to show motions and action and so forth. Maybe what we will see is an expansion of things like the indie manga artists in to indie anime and so on, a great subspace of artists and people who prefer human created content.
As always, I want to live to be 100 because it will be fun to watch…
“My personal ethos about sex is going to permeate the story and this is true of any author. ”
and
“Unless they do someday become sentient.”
If they do ever become sentient, then their personal ethos will come across. It will (likely!) be different from ours, and in that case, learning about our electronic children will be the point.
I’m not anti-capitalist. But I am pro-realist. I like to paraphrase Churchill: capitalism is the worst economic system on the planet, except for everything that came before it. I think you’re 100% right in how you characterize those who control the distribution and production. Harlan Ellison, in Dreams with Sharp Teeth (highly recommended, BTW), said that producers and those who finance a movie will be the writer’s best friend, right up until the story is delivered. Then the writer’s treated like trash.
I only understood that when I understood the same thing you articulated.
Why do those with power characterize creative people as lazy? So they can manipulate them. Turn public opinion against a group, and you can do pretty much what you want to them.
Why would you pay a lazy person anything? They’ll just keep being lazy.
I grew up on a farm. I’ve had farmers suggest I am lazy because I sit in an office building all day. They’ll point to diesel engine they’ve disassembled and say that’s work. That’s complicated. Those little beige boxes are nothing.
Years ago, that got to me. I bought into that ethos. I remembered coming in from the fields covered in dust and stumbling with exhaustion. I have it easy now, don’t it?
Sort of. Pros and cons. The Java/Apache Tomcat applications I write can have tens of thousands of lines of code. They rely on the open source Apache Tomcat to run, and it has hundreds of thousands of lines of code. A single browser in California has to traverse hundreds or thousands of miles of internet, with its routers and switches, to get to my servers. Some of them are sitting in AWS.
The bottom line? That enormous tractor engine sitting dissembled in the shop? It’s a fraction of the complexity of the software and hardware I’m utilizing.
Which is “real” work? Which deserves more accolates?
That’s the debate those in charge of distribution and production want us to have. It’s the wrong question. That tractor engine is critical to farming, and since I like to eat, I like farming.
My software powered wire transfers for credit unions, which serve those who are under-served in the financial industry. They can’t eat without money; my software helped them keep and better use their money.
Creative work is as important as any other work. I’d say it’s even more important, because it feeds the soul — if there’s another soul on the side creating the work. Otherwise, it’s popcorn — enjoyable, maybe, but without connection and without human meaning.
It does indeed feed the soul. It also feeds hundreds of people who work for publishing companies who don’t write – from the CEO to editors to the guy who runs the printing press machine. Yes, you are exactly correct – people doing hard manual labor think office people don’t work, office people think manual labor doesn’t work or is stupid – this whole idea of setting us up to compare and compete with each other for hardest worker – isn’t it the capitalists dream to have the workers so busy with that they never realize the CEO makes over 100x what they do and what the heck does s/he do? Never mind what the board makes, what the investors earn. I don’t actually know for sure – but I kind of wonder what the coders who create the AIs earn? Because they are creators, too.
And if our AI children become creators – how interesting that will be…
This is AI writed? “,”
Absolutely not!
But that’s just what you’d expect an AI to say, isn’t it? 🙂
You’ve expressed exactly one of the reasons I have no interest in AI “art”. I’m also not in favor of shutting AI tools down and think they can have great uses, but someone who writes prompts and feeds them into an image generator is certainly not an artists. They’re not expressing themselves but are generating an image that has a very general connection to something they might want to see. The connection between the human-created prompts and the end result are far too weak to call these people artists, and I believe that’s part of the reason the US Copyright Office has been denying copyright to prompt writers.
Since you bring up novels, I’ve noticed exactly the same in writing. There’s a new generative AI product now that can put together novels apparently, and supposedly this is meant to make it easier for the writer to write, to get past mental blocks, etc. But if I’m writing, I’m doing it to express myself, and I don’t intend to fill in even one paragraph with AI-generated content that doesn’t reflect my style or vision.
Sorry for the long comment, but I also feel strongly about this. Also as an enjoyer of anime girl pinup art who goes to Pixiv now and then. Thanks for the new artist suggestions!
No need to apologize at all — I appreciate you sharing your perspective!
I tend to think as you’ve described — namely, thinking that anyone who’s using an AI to write their novel isn’t writing a novel at all. They’re describing what they want to see. I’ve tried that a couple of times just to get a feel for it and to confirm it was as I thought. And it was. Nothing of me, except at the level I’d consider an elevator pitch, existing in that prose.
In such a novel, there’s nothing of who I am. No real insight into my perspective. So, there’s no communication.
Saying that, I don’t want to discount what DerekL referred to as democratizing. I don’t want to deny someone access to a technology that assists them. Am I being vain, for example, in thinking that the process of writing — coming up with the background, laying out the plot, writing each chapter word by word and sentence by sentence — is the only real way to write?
There are no new stories. AIs take advantage of that by sampling what’s come before. I acknowledge that there are no new stories. However, I will assert that there are new perspectives. An artist brings their experience with them to a story, and when they tell it, it’s that perspective that makes it worth reading.
I’m not denigrating anyone who writes just to sell; anyone who really doesn’t care to share their perspective. But I’m saying that’s not that I want to do. So maybe I’m saying someone who just wants to crank books out can do that if it suits them.
On the other hand, I’ll say this: as a writer, I look forward to competing with them.
“But if I’m writing, I’m doing it to express myself, and I don’t intend to fill in even one paragraph with AI-generated content that doesn’t reflect my style or vision.”
I’ll second that!
A couple things-
The AI might be including a much wider range of anime from outside the shonen demographic. If Sailor Moon, Cardcaptor Sakura, Ojamajo Doremi, plus all the romantic shojo series are in there it’s going to influence the body type. All the Precures are pretty slender, even the 17 and 18-year-olds, and that’s almost a thousand episodes and 80+ girls right there.
AI image generation varies a lot by what ‘seed’ you give it, and usually it isn’t clear how it arrives at what it produces. I suspect that asking for ‘girl’ might skew the results for anime, because the AI is ‘thinking’ of a much younger character for part of its input.
Case in point- I was playing with Midjourney to get ideas for what the magical helpers in my current work-in-progress would look like, and I used the term “fairy” because that’s what they’re all called in Precure, and I don’t particularly think of the fae as having wings. I was hoping for things like an anthropomorphic fox, but for “fox fairy” it gave me foxes with Tinkerbell wings pasted on. To get something remotely what I was hoping for I’ll need to do a lot of wordsmithing to push it in the right direction.
I was chatting with a couple AI researchers at Penguicon this year, and apparently if they can figure out what source(s) go into an AI response it’s an extreme rarity. Like infinitely nested black boxes.
“AI image generation varies a lot by what ‘seed’ you give it, and usually it isn’t clear how it arrives at what it produces.”
Yeah, my approach lacked precision. I traded depth for a single insight derived from a small sampling of images. I’m generalizing from a small input sample, and I know that’s not a great approach.
However, when I try to work the conclusion outward (art as communication doesn’t seem to work in direct proportion to the distances humans are from the product), it seems to hold true.
“Like infinitely nested black boxes.”
That’s a great way to describe it. At this level, my concern is that the AI is pulling patterns from works that, as you mentioned, would be difficult to cite. On one hand, that means the AI’s producer can claim ignorance. On the other, it means works are influencing the AI without attribution. That’s a concern for teh artist, isn’t it?
Humans do something similar. Nothing I write is independent from the culture that produced it. If I create a derivative work consciously, I should attribute it to the original artist, or to the artists who work I encountered. Some influences are subconscious — my opinion of the whole genre of high fantasy, for example, owes itself entirely to Tolkien. What I haven’t thought through is: what’s the difference between an AI sampling and combining and me incorporating cultural influences?
“But they weren’t the product of any human kinks – or any human affections.”
If the subject in an AI generated image has huge plot or a lush backstory or a winsome smile, or whatever… It’s because the artist has make a deliberate choice to include that feature. AI art isn’t simply pushing a button, there’s a human at the keyboard making conscious choices (good ones or bad ones) at every step of the process. (And that’s not considering the humans involved in programming and training the AI software.)
So how is an AI image with huge plot not reflective of the artist’s kink for huge plots? I mean sure, there’s those who create images that feature what they know is popular, but that’s true of all forms of art. (And deployed airbags aren’t exactly unique to AI assisted creators…) So, honestly, I can’t grasp where you’re coming from. Your distinction between an artist who draws huge plots on a blank screen and an artist who choses huge plots seems to be artificial at some levels.
Especially since my photography has elements of the latter case… I don’t create from a blank canvas, I choose from and work with what the world presents to me. The final image is the result of conscious decisions at each step in my workflow from raw image capture to finished image. Yet few would deny that I create art. (Not necessarily good art mind you! But art nonetheless.) And Photoshop, like all image programs, is also algorithmically based…
I can’t help but wonder if you’re reacting to the same thing we saw when smartphones with cameras democratized photography. The first thing is that the vast majority of cellphone photographers aren’t very good photographers… If their images turned out at all, it was a result of the algorithms embedded in the image processing firmware/hardware. The second is that cellphone photographers are very sharply limited by their tools – even as good as they are today, cell phone cameras are generally still inferior to a professional grade camera. (But a talented artist can still create amazing things… Which circles back to the first point. And the reason I use a third party photography app for my more serious cell phone work.)
Of the potential for subtle bias from the current social meta (“AI IS BAD”), I’ll just limit myself to mentioning the possibility.
Regarding your first paragraph, it’s the “training the AI” that concerns me. Yes, humans absorb artistic works, and yes, they often put their spin on it and create something new. That’s one way culture advances/changes/regresses. Thing is, the way AI works not, I’m uncomfortable with how little credit or attribution the AI routines give the source material.
I have no problem with the possibility that I have precisely zero idea of what I’m talking about. I just know — insofar as I can claim to know anything — that when I look at an AI-generated image (and by that, I mean an image that, as far as I can tell, is predominantly the result of AI-processed creation), my reaction is different than when I see a creation that’s predominantly from a human. I got this feeling as I described above, and on reflection, the description to match the facts.
That’s about the best I can do.
Am I reacting as a luddite? I doubt it. I embrace all new tech. I’m not even not embracing AI. I tried to outline where I think it would be helpful within my fields. And as far as me being influence by media, I suppose it’s possible. But I doubt it. That’s not too terribly consistent with how I view life.
I mean, Re:CREATORS is my favorite anime, for God’s sake. Does that sound like I even considered the Group Think’s opinion?
The concept of democratizing is interesting. In terms of writing or creation of graphic content, both take a certain level of skill — the question is, where do we draw the line between needing the skill and removing a specific human at a specific point in time from the equation?
For example, if I use ChatGPT or similar interface, or if I work with OpenAPI on my own, where does my vision for my art end and the AI’s skill begin? I’ve played with various scenarios, like describing a specific kind of space opera (very specific). The AI came up with a plausible, though bland, opening. If I try to publish that as is, is there any point? Anyone could have created that — and in fact, someones did (the way AI absorbs and converts content).
Let’s say I start adding paragraphs or changing some of the words. At what point is that more me than AI?
I come back to two reasons to create art in the first place:
1. As a means to communicate
2. As a way to drive “engagement”
Personally, I’m not interested in the second. I admitted in my post that I don’t object if people want to pursue the second. I think AI can play a support role in the first; AI could take over the second for all I care. I don’t want to read posts or look at art that’s not the product of a human’s act of will.
What I can’t answer is how much AI contribution is too much and how much is too little.
You likely haven’t read my novels, so you don’t know where I’m coming from when I talk about the philosophical and theological aspects of AI. Because if AI actually become sentient, then how I’d think about these equations would change dramatically.
I totally agree, AI can be used to help people in wonderful ways, from inspiring artists to crunching data for scientists, but it should never replace a human’s connection to another human.
I like that summary! Very clear and concise!
Absolutely this. I can see their use as a tool similar to how we use spellchecker of different filters on pictures. It can help us to tell our story better, but it cannot tell our story no matter how many prompts you throw at it.
And then there’s the legal ramifications of using someone’s work without their consent. It hasn’t started yet, but there will be a time when these AI tools are going to be forced to reveal how they trained them. In reality, it’s not AI, it’s just an algorithm that is putting patterns together based on large data sets. You could argue that we do the same without realizing it, but we can break the rules and patterns. We can experiment.
I have no plans of allowing an AI to do the best part of the work for me. Can it get me lunch or make me a cup of tea? That’s what I need.
I’m trying to find the right line — for example, I don’t want to limit what an AI can do on one hand. On the other, I don’t want to consume AI-centric work. What’s the balance? I’m not sure, but right now, based on what I understand of OpenAI and AI theory, based on the philosophical and (yes) theological aspects, I want to make a clear delineation between our respective roles.
I didn’t get into either the philosophical or theological aspects in this — I wanted simply to convey the idea that I think art should convey something of the artist, and then make it clear I have no issue with folks who want to create content without communicating. Though I’m not really interested in consuming that kind of content.
Now, if I could get an AI to control the machine to brew the perfect cup of tea? We might have the basis for a barter system between human and AI!
It’s a hot topic for sure and while I can see why people would use it to create stuff, I don’t think they can then say that they made it because they fed some prompts into a computer and watched it steal elements from other people’s actual art.
We grew up being told that computers would make our lives easier. Robots would do all the work that we don’t want to, allowing us the time to create and discover. Seems like it’s all backwards at the moment.