Mildly Infuriating

Using copyleft licenses for closed models is clearly against the spirit of the licenses if the users don’t have access to the source code that includes the original copyleft works. Even open weight models aren’t really the source code, and are more akin to a compiled binary. The source code is all the training data and code used to train the model such that anyone can build on it and train new models.

I’m not a lawyer and am not sure how well existing copyleft licenses like GPL or CC-SA would stand up in court to enforce this, but if they don’t, then stronger licenses that explicitly cover works being used as training data need to become more common.

I’ve seen the argument that the models are just learning from the data in the same way a human would. That’s nonsense. It’s not like they’re creating a sentient being with its own agency that can tell them to fuck off if it wants. These companies are running a software pipeline against copyrighted IP to convert it into a derivative work that is now supposedly wholly owned by said company, but the reality is that it’s collectively owned by everyone who contributed to the copyleft training data.

It's cheap. I don't have good enough hardware to run these models anyways.

Justice is not only about what is fair, but also about the fear people have of suffering consequences, aka: Pain. Being in jail is painful. It's a tiny room with no hygiene, cockroaches, the toilet is right there in the same room, isolation, unlikable neighbors, years or decades without Internet, a TV or PC...

The biggest lie the Jeffrey Epstein Class managed to pull off was to convince everyone that causing pain is an "absolute" wrong. It is not. Pain is what forces us to reflect and grow the fuck up to avoid committing the same mistakes. Those fuckers didn't have enough pain in their lives, because they lived it on Easy Mode. They were surrounded by yes-men and serfs constantly glazing them and laughing at their bad jokes. Now we are here.

We don't do shit to them because we are afraid of the pain the bodyguards, the cops or the law will lay upon us. But it should not be that way. They should be afraid of the people. Right now, they are laughing at us, doing heinous shit on purpose just to test our limit, and in the worst case scenario, normalize it. Young kids are already memeing Epstein, almost treating it like it's no big deal or just a joke, I seriously hope they don't carry that attitude forever, and realize that people like Epstein deserve to have their head opened to the truth as soon as you see them. Their bodyguards, I'd say are the worst people of all. The army, cops, bodyguards; they are people who openly and proudly decided to give their lives for a dude who rapes children and is not paying people what they're truly worth, a livable wage.

Nothing will get better until the rich fear for their lives.

They are also trying to take away these free resources by pushing laws to make ID identification mandatory to make money off of us.

The internet will never be open again because of this.

Once you understand that AI is limited to the questioner's ability to properly elucidate what they need to know you'll have several more botched concrete stair resurfacings.

Thanks Gemini, you self-contradicting potato

Apparently you had to sign an agreement that you will not use internet for business/commercial purposes in the NSFNET days.

https://en.wikipedia.org/wiki/National_Science_Foundation_Network#Acceptable_use_policy

20 years? More like somewhere between 30 to 40 if we count early WWW and the Gopher+Usenet that came before it. The GPL isn't quite that old, but the spirit behind it sure is. If we count early home-computer clubs back in the 70's (like the one that birthed the original Apple) or the ham-radio crowd that came before it, we can push into 50+ year territory, easy.

I hadn't considered AI being a paywall around the whole WWW but now that you mention it, it kind of looks that way. I've opined elsewhere that social media companies (e.g. Facebook) are building walled-gardens to keep eyeballs and attention-spans locked on their brand of reality. This would just be another avenue of attack in that strategy.

Welcome to the world of scientific publishing, long before AI. Except authors even have to pay for creating "content", and reviewers are expected to work for free. Yet article access is sold at astronomical prices.

That is how capitalism works.

The American Weather Service provided weather updates for free, but a company came along and just started copying what the weather service posted ... then sued the Weather Service for publicly posting the weather because the government is not allowed to provide a service for free that a company can charge for.

Insanity.

Coincidentally, OOP just explained academic publishing to a T.

There is a license that says that all derivatives must also be open source.

But also AI companies don't care about the law, they stole all their data, engage in insider trading, circular trading, and generating all manner of illegal content, they don't give a fuck. And the US government isn't doing anything to hold them accountable, infact the president is getting in on it.

I think this is the story of humanity. It ain't getting better until we massacre the inhumanly rich, eat Thier families while the world watches, and force evil socialism on shared intellectual property

Follow me for more bad advice

...Seriously?

This guy has no leg to stand on; that's the Mildly Infuriating part.

EDIT: Am I the only one on Lemmy who discounts the whole post the moment I see a blue checkmark?

Especially this one. It's so hypocritical it hurts.

They should be sued for theft and plagiarism.

The "you are replaceable" thing is older than everyone currently alive. It hasn't been invented by the AI tech bros.
Also, the library is still out there. You still can ignore the rent-seeking middle men and use it directly with your own natural neural network.

Not to mention those same companies obsessed with AI are the ones who run the search engines. They made finding all those tutorials and other good resources harder. They ruin search results with ads and easily gamed algorithms that they stopped trying to improve. All that made people more willing to let the AI find the answer.