Notices by yoasif (yoasif@mastodon.social)

@TechTangents @thomasfuchs What I am confused about is why fair use has been expanded to encompass wholesale piracy - does copyright law allow someone to abuse fair use to "transform" the original work into a virtual facsimile?

It seems to me that somehow, the theory of the case is that if I can copy a sentence, it is fair use, so I can copy a paragraph, and a chapter, and then a book.

Humans can't do that, but LLMs can.

Is there some nuance we are missing or is this a new legal theory?

@hagen Worth boosting the struggle of @wwu in its fight for unionizing the #wikimedia foundation.

https://www.quippd.com/writing/2026/08/04/wikipedia-turned-on-its-volunteers-now-its-coming-for-its-workers.html

@MegaMichelle @jbowen @sarahjamielewis @servo It definitely makes sense to build an all new codebase instead of forking an existing one while the new one matures. 🙄

What are we supposed to do until Servo is ready?

@inthehands I think the one big tech is telling themselves is that companies will replace their labor costs with token spend -- and they will pay what it costs.

Deskilling is part of it, which is absolutely happening, and for some people, they never gained the skill - so without the labor availability, you replace a human staff with tokens.

That's the dream anyway.

The bigger question is that even if the bubble pops, how long will the well remain poisoned - will we ever go back to sharing?

@dalias @ariadne I said different, but I suppose I assumed that the GPL projects @ariadne was thinking about the ones who have given into slop coding; GPL code that isn't GPL, essentially.

I think if people are the ones developing the codebase, we are still stuck with the problem of derivative works not complying with the GPL (as the slop coded projects are derivatives of the GPL projects that have been trained on).

Copyleft becomes less of a cudgel - I don't think it is anything but weaker.

@darnell @jalefkowit @ricmac About using it to check code... https://www.backslash.security/blog/can-ai-vibe-coding-be-trusted

The same applies to everything else, of course.

@weebdeluxe I wrote a little about #Mozilla pivot to #AI as well: https://www.quippd.com/writing/2026/02/09/firefoxs-ai-kill-switch-is-a-trap-how-mozilla-made-ai-your-problem.html

You may enjoy it.

I have largely moved to @librewolf FWIW.

@dalias @quarknova I'm talking about contributing - this whole thread is about contributing.

If you are talking about something else, I'm happy to discontinue discussing.

@dalias @quarknova I don't agree that the commons is preserved, FWIW - I think continuing to contribute acknowledges that your contributions are given as a kind of dual-license - share-alike for the commons, proprietary (and paid for) for big tech.

Downstream from that, since the license is ignored when the trained LLMs output the contributions as public domain, you have preserved the contribution in a very different way than intended -- kind of a monkey paw effect.

@dalias @quarknova I agree - I am simply trying to make people aware of the fact that your contributions are no longer protected by share-alike when contributing to Wikipedia, since Wikipedia has opted you out.

If you don't think your contributions should be helping drive the slop that we all bemoan, it is probably a good idea to not continue to fund it for free.

Big Tech can hire people to edit a Wikipedia-like corpus that could conceivably better than Wikipedia; they've got the money.

@dalias @quarknova I mean, forks could exist.

Forks wouldn't have a deal for big tech to opt out of share-alike, so we could conceivably see better stewards for this repository of knowledge than the Jimmy Wales crew.

As it is, contributing directly to Wikipedia constitutes a tacit acknowledgement of a rejection of the license that your contributions are nominally contributed under; Wikipedia has opted you out, even if you haven't.

https://www.youtube.com/watch?v=K0omu7x_LbU

@dalias @quarknova I don't actually believe that the bots can generate knowledge, but that is what we are told.

The lie is revealed with the demand for real human knowledge.

It isn't "malevolent capitalists" -- it is Wikipedia that has sold out its community: https://www.avclub.com/wikipedia-ai-partnerships-meta-amazon-microsoft

If it is good for the goose, why isn't good enough for the gander?

More precisely, why continue to contribute to Wikipedia when they have preemptively sold out the commons?

@quarknova I think that unfortunately, this policy is exactly what big tech wants - fresh training knowledge and training data from humans, which the tech companies are free to remix into new public domain works that are not protected by share-alike.

With this policy, Wikipedia demands that humans work for the bots for free, and the humans can't even take advantage of the bots to work against the theft of the commons.

@inthehands @julieofthespirits US WW2 - a better example of a concentration camp that wasn't a death camp.

@inthehands @julieofthespirits The internment camps were concentration camps, so that might be a better example.

@joepie91 @firefoxwebdevs Mozilla's tortured definition of opt-in seems to predict that Mozilla will invent features to nag you into enabling AI, as they have already done with Link Previews: https://www.quippd.com/writing/2026/01/06/architecting-consent-for-ai-deceptive-patterns-in-firefox-link-previews.html

Is #Mozilla new CEO actually threatening the removal of MV2 based ad blockers in #Firefox on his FIRST DAY as CEO?

"He says he could begin to block ad blockers in Firefox and estimates that’d bring in another $150 million, but he doesn’t want to do that. It feels off-mission."

Why is this even on your mind?

🤯

https://www.theverge.com/tech/845216/mozilla-ceo-anthony-enzor-demeo

@killyourfm #Mozilla is moving their support pages to AI translations: https://www.quippd.com/writing/2025/11/27/thank-mozilla-for-killing-localization-on-support-mozilla-and-replacing-human-contributions-with-AI-bots.html

😉

@skinnylatte Will need to visit soon!

@skinnylatte Where?