If a single person can publish 200 novels a year, they drown the market in low quality contents. The payback per unit of work (book) becomes insignificant, such that no new actor (other than those also able to publish 200 books a year) and no new ideas (that requires actual work) can enter. An art is destroyed, a market is destroyed, a livelihood (for many actors) is destroyed. It's bad for everyone except for 1 person.

The company cited by the OP associated belonging to Eurovision as a sign of association to the EU. This lead me to remind that for example Brazil and Japan have the same level of membership as Israel in Eurovision's parent organisation.

It's just a numbering scheme. Linus thinks .19 is already big enough number to increment the major release number, even without any new feature introduction. So 6.19 is the last 6.x version, as had happened with 3.19, 4.20, 5.19.

The argument of the AC is code sharing. For example, the Stellantis group might reuse parts of their code between Jeep and Fiat. If a mole, a suitable EO from a stakeholder government (the USA or Italy in this example), or an external bad actor, get remote access to the cars of one brand through a rootkit or a vulnerability, they could also simultaneously gain access to other other brands.

Eurovision is not an EU organ, it is the information exchange network of the European Broadcasting Union (EBU), created in 1950, previous to any of the predecessors of the current EU. Its members are *broadcasters*, not countries; but reach to the USA, Brazil, Chile, Japan, China, Malaysia, Morocco, South Africa, New Zealand, etc. One of the achievements of the EBU was the development of technology enabling simultaneous broadcasting. As a demonstrator to their broadcast technology, the EBU started a song contest in 1954. Programs shown in Eurovision network mode start with the Anthem of the EBU, based on the first notes of Marc-Antoine Charpentier's Te Deum.

It might surprise you, but that's a common thing. People get interested in books, but then don't make up time for reading. A long time ago as graduate student I did not have internet at home, I was purchasing *and* reading a book every week, whether an essay or a novel. Now I have internet and also a busy life, and I can't make up time for reading anymore. I probably could, but there so many other things I can push forward even on nights and week-ends: my main job, my side hustle, a hobby (e.g. street photography), contributing to some FOSS, trying to play an instrument again, playing with the cat, or, you know, waste time online.

Here the first lines of another of his prose. He admits to writing stupid things online all the time, to the point he does not give his real name in job interviews. https://www.experimental-histo...

What's TSO? The whole first page of google results for "TSO main screen" is IBM z/OS. Wikipedia has many meanings like "Tokyo Symphony Orchestra" and other things that seem equally irrelevant https://en.wikipedia.org/wiki/...

What's your problem with the AI features? There's nothing on a Firefox page that makes an AI feature apparent if you don't trigger them in a menu.

The "translation" icon only shows up as an icon when you visit a page which declares a language out of your http-accept-languages settings (at which point you actually need a translation). There is also a "ask a chatbot" option in the right click menu, and it's doing nothing if you don't you configure account at a chatbot of your choice.

I really don't understand what you think is wrong with Mozilla's approach. If you don't like AI features, just don't use them. They're buried in a menu, you won't even know they exist.

"did everything well from day 1"

Yeah, the translation feature is constantly helpful for people living in multilingual environments (the model training and FF integration were funded by the EU), and generally people not native in English everywhere. It has helped me with technical documentation available in Chinese, and it's a convenient way to have access to certain topics of Wikipedia without leaking my interests to Google.

I don't get ethe anti-AI hate wrt Firefox. Mozilla does not host AI cloud services, FF uses a local model, what is not local is opt-in since an external account with a chatbot is needed. Mozilla did everything from day 1.

Spellcheck isn't "generative AI", and does not make the content "substantially authored" by the spellchecker.
The scope of the law, from TFA: "substantially composed, authored, or created through the use of generative artificial intelligence.”

My arguments in reply of your concerns: 1) Definition: They don't have to define AI in the law; the newspaper are not going to invent some new mathematical approach, they are going to use commercial tools. The newspapers will self-incriminate by using tools that claim to be AI. Anyway LLM is another worg they can include.
2) Proof in court: It is enforceable as any regulation of a process that isn't immediately visible (like in the food or water industry). A whistleblower will talk and bring internal emails.
3) Constitution: I can't tell myself, but when it reaches that far in the legislative process, it has been reviewed by people who know a lot.

Came here to say that. They should have used generative AI for things like crowds, scenery, monsters, where AI excels and nobody can tell if it's slightly off. AI is great in cutting the cost of something like LOTR. But human faces... the slightest detail off and it becomes a nightmare character; very unwise to start there.

If you want a vehicle that does not use data-related features, you don't have to rule out BMW, just just have to not pay their subscription. (It's even the point of this story.)