Micro and small enterprises are excepted from some rules but far from all.

I think that lemmy servers would count individually, as well, but it is not guaranteed. In any case, that comes with its own problems.

That’s not correct. The FT has not explained this clearly.

If an online platform has more than 45 million monthly users (~10% of EU population) then it is classified as a Very Large Online Platform. In that case, the Commission can directly make rules for it.

If it has fewer users, then it is still regulated by the Digital Services Act (DSA). The DSA claims jurisdiction over all platforms that have users in the EU. Among other things, they need to have a representative in the EU (IIUC). FWIW I’m pretty sure that lemmy is not compliant either.

DSA: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32022R2065&qid=1732567528372

The article is fake news. I suggest looking elsewhere for proper information.

As for your questions: LLMs were certainly not involved here. I can’t guess what techniques were used.

Racial discrimination is often hard to nail down. Race is implicit in any number of facts. Place of birth, current address, school, … You could infer race from such data. If you do not look at race at all but the end result still discriminates, then it’s probably still racial discrimination. I say probably because you are free to do what you like and discriminate based on any number of factors, as long as it isn’t race, sex, and the like. You certainly may discriminate based on education or wealth. Things being as they are, that will discriminate against minorities. They have systematically lower credit ratings, for example.

In the case of generative AI, bias is often not clearly defined. For example, you type “US President” into an image generator. All US presidents so far were male, and all but one white. But half of all people who are eligible for the presidency are female and (I think) a little less than half non-white. So what’s the non-biased output?

Hmm. It was a big issue at the time. In truth, I’m really not sure how it works in France. Anyway, the big fight going on is really about minimal previews. Unfortunately, there is no disinterested reporting on the issue. The media is very much profit-maximizing.

The recitals aren’t part of the law, but should only guide the interpretation. Also, this is a directive. That means it directs the member states to make law, but has no direct effect, as such.

My brother in Lemmy, this is what stopping Big Tech looks like.

Europe made laws that say that Google and others need to pay if they want to link to EU publishers. Well, maybe the price they are asking is not worth it.

You’re right about the firewall energy, but that’s simply how these laws work. The point of copyright, as well as age verification and other such laws, is to control who may access certain information.

Because that’s one key feature in the “2019 European directive adopted into French law”. It’s also what the Google fine was about.

Also, X isn’t really suitable for copy/pasting entire articles, like is done on lemmy. So that’s probably not it.

Misunderstanding. The news media is suing X. I pointed out what news media does without paying.

You think people should pay X to link to tweets? Or generally for quotes?

More like: They want to sell the cake and be paid when you recommend it to others.

Mind that news media don’t pay when they link to social media, quote people, or even report what other media has reported. The real question is, if this law has any beneficial effect for society. I don’t see how.

Oh, he’s saying that snippet view lets us have sites like lemmy. I didn’t get how cracking down on that would help lemmy.

It should only show the title and the link imo.

That’s infringement in Europe, which makes it effectively a link tax.

Huh? How you mean?

Not quite. You can’t turn movies into books or games, or vice versa, for example. Sometimes such projects get stuck in limbo. Or think about how everyone hated the final season of Game of Thrones. Can’t do anything about that in our life times.

The “battle” is the result of copyright people trying to use open source people for their ends.

In the past, for software, the focus was completely on the terms of the license. If you look at OSI’s new definition, you will find no mention of that, despite the fact that common licenses in the AI world are not in line with traditional standards. The big focus is data, because that is what copyright people care about. AI trainers are supposed to provide extensive documentation on training data. That’s exactly the same demand that the copyright lobby managed to get into the european AI Act. They will use that to sue people for piracy.

Of course, what the copyright people really want is free money. They’re spreading the myth that training data is like source code and training like compiling. That may seem like a harmless, flawed analogy. But the implication is that the people who work and pay to do open source AI have actually done nothing except piracy. If they can convince judges or politicians who don’t understand the implications then this may cause a lot of damage.

Other way around. The NNs are written in, mostly, Python. The frameworks, mainly Pytorch now, handle the heavy-duty math.

Musk also has his hands on a fair amount of data through X and Tesla. But yeah… Copyright expansion seems like an odd place to start breaking the constitution.

You’re right about the regulation but I’m not so sure about the copyright exemptions. All in all, you’d think he’s more likely to side with property owners - especially the heirs of media empires - over progress.

Why is she claiming that the bill is about liability?

The winners of a system don’t have an incentive to undermine the rules. Quite the opposite. The NYT wants these rules because it would benefit from them. There are at least 2 image generators that adhere to capitalist ethics. I don’t know what Claro uses, but I see no indication that they are being uppity.

You are apparently mistaking me for someone else.