zlacker

>>dend+(OP)
> Exploiting user-generated content.

You know, if I've noticed anything in the past couple years, it's that even if you self-host your own site, it's still going to get hoovered up and used/exploited by things like AI training bots. I think between everyone's code getting trained on, even if it's AGPLv3 or something similarly restrictive, and generally everything public on the internet getting "trained" and "transformed" to basically launder it via "AI", I can absolutely see why someone rational would want to share a whole lot less, anywhere, in an open fashion, regardless of where it's hosted.

I'd honestly rather see and think more about how to segment communities locally, and go back to the "fragmented" way things once were. It's easier to want to share with other real people than inadvertently working for free to enrich companies.

>>rpcope+9j
You know, if I've noticed anything in the past couple years, it's that even if you self-host your own site, it's still going to get hoovered up and used/exploited by things like AI training bots.

So? What do I care? If some stuff I posted to my website (with no requirement for attribution or remuneration, and also no guarantee that the information is true or valid) can improve the AI services that I use, great.

>>Camper+dl
I think the source of the contrary sentiment goes something like this: AI stuff (especially image generation) is competition for artists. They don't much like competition that can easily undercut them on price, so they want to veto it somehow and lean on their go-to of accusing anybody who competes with them of theft.

The problem in this case is that it doesn't matter. The AI stuff is going to exist, and compete with them, whether the AI companies have to pay some pittance for training data or not.

But the chorus is made worse by two major factors.

First, many of the AI companies themselves are closed-source profiteers. "OpenAI" stepping all over themselves to be the opposite of their own name etc. If all the models got trained and then published, people would be much more inclined to say "oh, this is neat, I can use this myself and it knows my own work". But when you have companies hoovering everything up for free and then trying to keep the result proprietary, they look like scumbags and that pisses people off.

Second, then you get other opportunistic scumbags who try to turn that legitimate ire into their own profit by claiming that training for free should be prohibited so that only proprietary models can be created.

Whereas the solution you actually want is that anybody can train a model on public data but then they have to publish the model/weights. Which is probably not going to happen because in practice the law is likely to end up being what favors one of the scumbags.