It has become fashionable on Threads and social media in general to have flame wars about AI in art, in writing, or in general, and the opinions are often quite wild. “AI is theft. AI is wrecking the planet. AI is techno fascism” . These are some of the refrains I hear repeatedly, and it’s quite interesting at times to listen to how stubbornly these views are held.
Anybody who happens to be using AI is fondly referred to as a slopper, or a crank wanker, or various other phrases. Humanity has a grand talent for creating rude labels. This is something I suspect AI may never beat us at. And I see these opinions. Sometimes I react. Sometimes I react well. Sometimes I react less well. Because I like language, rhetoric, and a reasonably well-formed sentence, I’ve had more than one smug idiot accuse me, determinedly, of being a bot, as apparently this is an insult as well as a bad thing and only bots are capable of well-formed sentences and half-decent rhetoric.
In one domain, I vastly prefer my own voice to that of an AI: I like writing at times. For fiction, I don’t really ever want to see an AI write for me, so it will not.
The blog posts on this site are a bit of a mix because many of them are not fiction. Many of them are log write-ups from source code. Here, you will see some co-authored stuff. I do intend to replace it because I still feel I’d rather it was my voice than Claude’s. This just takes time and motivation.
Let’s address theft. I’ve repeated this argument many, many times: you break into a gallery and don’t pay for entry. You’ve committed trespass. You don’t steal or destroy any paintings. You study them for hours, you make notes, you take videos. Perhaps you do this more than once. Perhaps you spend a year doing it, and perhaps you do it to more than one gallery. Perhaps you borrow or steal a lot of books, and then you read them carefully, slowly. You take apart the techniques, and you become an expert in the techniques you’ve taken from these books and these paintings.
Now you write your own books, you paint your own paintings, and you use everything you’ve ever learned to produce them. 99.9999% of the time, you don’t directly, word for word, line for line, reproduce somebody else’s work, though you often ape a style. What are you guilty of? The original trespass and the original theft, they still stand. You didn’t pay your entry fee. You did steal a book. Now you owe somebody money for that book. That’s it.
The only other thing you’re guilty of, potentially, is being cheap hack, if you really have ape’d another artist to the point where their style is quite recognisable, and you get to be pilloried for that if it’s too obvious. The work you produce then doesn’t become particularly valuable, at least to most. That’s as far as it really goes.
Pretty much, most art and most writing is influenced by other artists and other writers in some way or other. The only way for that to not be true is to have been brought up in an isolated cave, many miles underground, away from all human contact. Your work isn’t theft. It may not always have good taste. It may not always be particularly original, but it’s not theft, regardless of how loud various people get. Loudness doesn’t equal truth.
As I write, I find myself reflecting on the fact that here and there I like rhetorical patterns. Rhetorical patterns, however, have come under fire. Under fire, because AI tends to overuse them, and overuse them quite often in a way that’s not entirely appropriate, or at least seems very heavy-handed. Heavy-handed use of rhetorical patterns does show itself and tends to make the mind weary, and a weary mind turns adrift off the page. I’ve done it myself, right here, quite deliberately. AI can be this clumsy too.
There’s no way for a writer to prove, beyond all doubt, that what they wrote isn’t AI. There’s no snake oil that you can use to prove it beyond all doubt or disprove it. It’s an ongoing arms race where companies promote their specific snake oil, their mathematical bandaid and distressingly, a certain number of institutes and companies have even tried to rely on these modern takes on numerology and phrenology to assess writing and its provenance.
They’re not actually going to succeed. They’re just going to think they are, because apparently a 1-in-10 failure is fine. When the numbers start getting above 100, this is really not the case, but never mind. So let’s move on to another piece of this debate: talent, creativity, hard work. This is a much more subjective area and doesn’t have nice, clean answers, unlike the AI theft debate.
Should all art be a result of your own manual hard work? What then of photographers, or anybody who uses any form of algorithm, anybody who uses any kind of mechanical assistance whatsoever? Where do we draw this line? Is the offensive part the fact that it seems too easy, that you just type in a prompt?
There isn’t a single answer, and each person’s response is going to be individual to them, depending on whether they’ve fought for years to develop their own skills or not. The strength of bias, I suspect, will be intimately tied to the depth of their own investment in their art.
This strikes me as a natural response. I began this little section with no specific hard opinion in mind. I do have a rebellious nature against gatekeeping of any form and being told that I must go along with a particular label, a particular category, or,being told, “You are not X if you don’t do Y, and you are Y if you don’t do X,” and all variations I see online. It does tend to spark an anti-authoritarian flame within me, even though it rarely even applies to me because I would not claim the title of artist myself.
Actually, I do find the notion of labelling somewhat problematic anyway, because people are rarely one identity, one thing, they are acting a role for a time. Sometimes that role is very dedicated. Sometimes it’s their entire life, but people change. A person is actually a process. The act of being conscious is a process. Living a life is a process. We are not frozen weights in a model. We are not statues. We always change. Change is what life is.
I could do this little philosophical ramble all morning long. However, I think art is something you do. I don’t think it’s a label or title you wear or can be gifted by others. I would imagine I am a complete minority on this specific point, but I don’t really care. To me, art is something you actually do, and you’re doing it with intention. The intention is usually to gain a reaction, to provoke a response, and communicate something. What you might be trying to communicate may not always be particularly clear to you. It may just be something deep inside you want to get out, and you don’t necessarily have all the words to explain why. Or what? Sometimes you do, sometimes you do not. Sometimes what comes out is what needs to come out, and even to you as the art producer, there may not be a clear idea why and what.
For me as a probable non-artist, I feel that art is something you do. It’s not a label you wear.
I suppose I should have a go at addressing talent and see how many brick bats I can possibly acquire in the process. That’s always good for a laugh.
Apparently, you’re born with it. Apparently, there’s a gigantic genetic lottery that flips switches and patterns in your brain, your muscles, and your bones, and gives you a gift beyond others of some description. Mozart is often cited this way, so are many others, but this is circumstantial evidence at best. We’re not inside their heads. We are not inside their memories, and we have no idea of what small little things they did inside their minds, time after time, obsessively, to gain a particular skill.
Now this is going to sound foreign to most people, but the notion of luck does enter the equation, and I think quite significantly. I believe it’s entirely possible for a very young mind to happen upon a pattern of thinking and a sort of self-awareness about something specific, often quite a narrow domain, that they don’t let go of. They keep chasing. A classic one would be simply that a child decided to keep counting backwards, or counting backwards in threes and fours. Or perhaps they tap their fingers in a sequence, (I did this for a short time) and they keep trying different sequences, for whatever reason.
Human brains have a habit of fixating. It’s often not necessarily terribly long or terribly deep when young, but at times it can happen. Whatever it fixates on, it learns really, really well. If it happens by chance to fixate on something that’s quite transferable, it’s acquired a new skill, a skill that’s above average compared to its peers.
I actually think an awful lot of talent can be explained by a confluence of such little obsessions and interests that occur often before they’re necessarily speaking a lot. Some of this is pure speculation. I just happen to think it’s more likely than somehow a magic genetic lottery was won in the brain, whereby one acquires the ability to be a genius in some particular field by the age of 5 and get a degree by the age of 10, which has apparently occurred.
I think it’s a huge amount of luck, but not necessarily some genetic talent. I’ll admit my bias here: I don’t like games of chance very much. As games of chance, on the whole, don’t truly require skill, at best an intuitive grasp of odds is all you’re going to have. If a game doesn’t have much you can influence via your own agency and skill, I fail to see the point. That’s my bias.
A game like talent, again, is pure chance. Of course reality doesn’t necessarily conform to our biases. This just merely explains why I’m biased this way. I have reasoning. I don’t think we have empirical evidence that talent, as is commonly used, is a real thing. I think we have a tonne of circumstantial evidence, a tonne of anecdotal evidence, a lot of romantic stories, and some startling successes in science and maths and music (that are, again, part of the circumstantial and anecdotal branch of evidence).
But we are not in those people’s heads. We don’t know what they did inside to become aware of the things that they thought of and then produced into the world. I actually feel that the term talent even cheapens achievement. I’m not alone in this. It relegates the credit to the talent first and the hard work second.
Though of course, if you substitute the word talent for luck, which, in some ways, I think perhaps you could, where the luck was, you happened to think in a certain way, at a certain time, enough times to develop a skill and a direction of thinking that led you to where you were (i.e., thinking of the special theory and general theories of relativity by apparently mentally riding a light beam around the universe).
Now of course, we all know that’s sort of a romantic simplification of what Einstein really did to get there. Though it wouldn’t surprise me to find that a series of little imaginings was a huge part of what led him down the line of all the hard work he would have had to do to formulate and formalise his theory.
This post has drifted off the topic very slightly because the main objection I’m seeing with AI use and creativity and talent is that, apparently, using AI means you are stripped of such creativity and apparent talent. I find it hard to feel that somebody did work as hard at a piece of work if they just used AI for it. I know that part, at least, is true.
If I prompt an AI for a winged super yacht, which I have done, and others have, the only original thing in the entire construct was my prompt. That didn’t exactly take a lot of brain power so I can fully understand why folks who work hard at their art would be somewhat disdainful of this process.
Oddly enough I don’t believe it makes it not art. I do believe it probably makes it cheap and I do believe it doesn’t necessarily showcase a person’s skill particularly. If you want to showcase your skill then you have to use your skill. If you just want to showcase some thinking, that’s actually a different prospect.
I think one thing we might want to consider in this debate is dropping some of the labelling aspects, as I pointed out earlier. I think this fixed notion that people take a title, artist, etc., is a bit outdated. Perhaps it works well for being paid for a job where you go to work as an artist and that’s how you earn your living. Beyond that I think it’s something you do.
If folks focused their debate on that, then you could fairly accuse somebody who only spent 10 minutes producing what should have been 2 days of work as not really doing very much actual art. They literally did spend 10 minutes to make it happen instead of spending 2 days engaged with the process of doing art. That isn’t gatekeeping. That’s just stating facts and one is definitely cognitively cheaper to do than the other. Whether the market then wants to pay somebody for either piece of work is a decision that’s entirely little to do with the current trends on social media.