I've got a dude arguing in the comments on one of my posts that he has to use AI to produce music because he doesn't have the money to do anything else, and that if we really think he shouldn't be using AI, we should give him some money to hire musicians.

...

This gets into some territory that I feel like I gotta dissect.

First, musicians /should/ get paid for their work.

Second, GenAI "music" more or less guarantees that the musicians who's work was stolen to produce that "music" are not getting paid.

Third, "Generative AI' is such an overloaded term that it's easy for lots of things to get wrapped up in the bucket that mean different things, and that makes talking about it difficult.

I don't want this to be a big thread, I don't want to be talking about Generative AI at all, but I've been ragebaited, so let's chat.

So, first, I suspended the person in question from this server, because he chose to follow me (after sending a dozen or more comments) *after* I asked him to leave me out of his GenAI evangelizing.

He called my stance against AI virtue signalling, so I figured we could talk about that.

When Generative AI hit the scene, it took me a really long time to make up my mind about it. Some people around me were (rightly) immediately against it, but I didn't fully understand why.

I played with early image generators because I found the results amusing. I was running stable diffusion locally, because I had a machine that could do that. I tried to convince myself that there were places that this tech could be used ethically. I tried to justify it.

I tinkered with early LLMs, partially because I found the results amusing and partially because (like many other people in the tech world) I saw in them the promise of a thing that I had long dreamed about and couldn't really articulate. (I was wrong! That promise is not there. These are machines built to look useful while doing nothing.)

And, in the meantime, I learned a lot about how these things work.

I've seen inside the sausage, I've eaten the sausage. I can say without question, that the sausage is made of people and sewage, and we're just hurting one another and ourselves when we eat it.

But there's nuance here, and I want to get into that for a second.

There are a couple of places that I still use or have used or plan to use things that are commonly rolled up in the "AI" bucket, and that I believe are fine.

Because "AI" is a bunch of different technologies built a bunch of different ways, wearing a marketing trenchcoat.

I use an automatic subtitle generation tool that is "AI" powered, but runs locally and isn't generating anything new, it's trying to identify a thing that already exists.

The argument could be made that it was trained on stolen labor in that it was trained by matching existing subtitles to the audio tracks those subtitles correspond to. I decided I was personally comfortable with this, but I understand how other people might not be.

It's wrong a lot, but it gives me a place to start from as I'm manually subtitling a thing, and saves me hours of transcription work.

I think it's fine. I think it's good even, I like it and I want it to get better.

I also occasionally use image classifiers to sort through large volumes of photos or scans or whatever because going through them by hand would take a lifetime, and some large corporations already taught computers to do this. This makes some of my research and art projects much faster, and can again be run locally and generally doesn't involve any stolen labor (at least no more than we've been giving away to solve captchas)

I'm planning to set up a particular image classifier that is designed to specifically extract foreground images from solid color backgrounds, to provide a clean chromakey for footage that was not filmed against a properly lit green screen.

Again, it's not generating anything new, it's automating manual labor away, and letting me focus on the creative part of video production, rather than rotoscoping a person out of their background over and over again.

I think this is great!

I don't think it's really fair to call either of these things "AI" in the way that we talk about "AI" right now, but they are both built on machine learning, and ...

I have a point, give me a second.