There are a lot of takes about AI out there on the internets, some good, some bad. Okay, actually, a lot bad. There’s talks of the environmental impact, which I personally feel can’t be spoken about fully, until we acknowledge the environmental impact that we, and our habits, have already contributed to environmental breakdown (fossil fuels, media consumption, consumerism). And then, there’s talks of copyright, and LLM training, which … for some reason, is continually popping people into kernels of a Lars Ulrich meme from 2000.
In discussions about LLM training, to many, it’s either theft, or it’s learning. Learning, in the way that a person might learn if they take art classes, or study art, or what-have-you. Wherein they learn about other artists, they learn from their works. Some people even adapt styles from others in order to incorporate said styles into their eventual own works. And, in some cases, you can compare this to how LLMs are trained.
Human beings learn by studying the art of others, on a very small scale, a little bit at a time. You could of course, exist in a vacuum your entire life, and be handed a pencil and a piece of paper, but I believe even in that environment, you would take cues from things around you. Even if those things are sparse, and almost nothing exists.
An LLM learns by being shown the work of others, at an extremely large scale, as if you’re taking the work of thousands of people and shoving it into your brain.
Either way, you could say that these are similar ways of learning, unless you’re of the assumption that an LLM cannot create anything new, which is arguable. It’s very arguable, if you’re making these statements, and you’ve never actually interacted with an LLM or generative media.
It’s also arguable that a person cannot create anything new, because every idea has already been done, and everything currently being done is just something being done in a different style, or a different theme, looked at from another angle, with different purpose, and that everything can be traced back to an origin, a source.
Take the Backrooms for example. An idea that originated from a 4chan forum thread? Or a concept that was born on Tumblr from someone who was talking about dead gas stations in the middle of the night on the side of a highway? Or, the 1990 David Lynch series, Twin Peaks?
And so on and so forth, you get the idea.
It’s the idea of training and learning being theft that brings me to the point of this blog post today. Wherein, the belief that training an LLM on the works of creatives around the world, has suddenly catapulted people into this idea that, if you’re a person reading or consuming art and media without a license, you’re stealing, and that’s bad.
It shouldn’t have to be said, but copyright is not good. The only people who benefit from copyright, are the rich, the billionaires. And in this example of Lars Ulrich, we had/have one of the wealthiest bands in the world, who went on a campaign in the early 2000s in order to rally against Napster, because people were downloading their music. This essentially killed Napster, and controversially is the tipping point that brought us to where we are today, where the mainstream way to consume music is to stream it.
When was the last time you saw a store that exclusively just sells CDs?
But, it’s due to the aforementioned thinking, that if someone downloads your album, that’s a lost sale.
But here’s the thing: If someone is pirating music, or art, or a game, they weren’t potential customers. In some cases, they might become customers, but it’s never an additional lost sale. Most people pirate because they can’t afford to buy, and I definitely wouldn’t see many issues with that today, when games are reaching 70 USD and more, and most don’t even own the music they listen to, anymore.
With this in mind, you can take this concept and apply it to generative media. If someone generates an image of an anime girl, I don’t know, punching someone, that’s not a lost commission that some artist isn’t making. That person who generated that image was never going to pay anyone a commission for art, whether AI exists or not.
Furthermore, AI isn’t stopping anyone from continuing to draw, to create, to compose, and to write.
Whether people like it or not, and how valid or invalid it is, the main point not being spoken about here, is competition. People perceive AI as competition, even if it isn’t. Even if piracy isn’t losing anyone any customers, or gen-AI losing anyone any commissions. And to rally around the concept of copyright, to the point where you’d say something like this:
I do see it differently. Looking at something or reading something isn't stealing unless you don't have the license to look at it or read it. Buying a book and reading it, and then being inspired by the content is actually good.
— Girl Lich 🏳️⚧️⚢💀 (@girllich.bsky.social) April 28, 2026 at 1:14 PM
… is probably one of the largest faux pas happening in online leftist spaces right now.
“You can’t look at my work unless you’ve paid for it.”
Look, sweaty, I know we’re all hurting for money and this economy is crushing us, and something must be done about it before it destroys the entire world, but seriously? You’re just going to say that? No irony at all?
How are we allowing ourselves to arrive at these dystopian mindsets, just because of AI/LLMs?
Maybe these are ideas that people have always held, and generative media is only bringing it to light, now, because it’s suddenly more acceptable to sound like a lawyer.
Anyway, on a side note, I’ve been speaking about the nuances of AI a bit more lately, and recently, as of a few days ago, someone who came to my defense during a bunch of harassment on the Fediverse (which I am still eternally grateful for, since nobody usually speaks up for me when these things happen), decided to then, and now, stab me in the back. Because anything but complete angry and vicious rejection of anything AI at all, is evil and bad. That because I’d even entertain these ideas, every single thing I do and post must also be AI generated, which is insulting for multiple reasons, even if there are things I do that have AI influence, or are generated by AI. For reference, I write all of my blog posts, and I pass them through an AI to generate tags and summaries for the metadata. You know, some of the simple monotonous stuff I just don’t feel like doing (see: Groq, not to be confused with the Elon Musk machine).
Yes, I am aware that I am essentially training some LLMs with the things I write. I also don’t care. As we’ve all seen before, having access to AI doesn’t mean you can produce good things, or, having access to AI doesn’t suddenly make you intelligent, or magically give you an imagination.
But I shouldn’t have to explain myself. What I do is my business, and one of the things I’ve learned over the past two decades, is that, generally, almost nobody online has any effect on my actual life. There are very few people who do influence me, and who care about my existence online, and to those people, I’m grateful they exist.
But, while I continue to deal with the superficiality of so-called good people in online spaces, maybe there’s something for more reasonable people to take away from this post.
Posted via Python, written by @cmdr_nova
