Notices by Luis Villa (luis_in_brief@social.coop)

RE: https://mastodon.social/@bd808/116133511257881088

I can’t believe it’s been a decade. Anyway here’s a post that’s not about this but it’s not not about this: https://lu.is/2021/04/values-centered-npos-with-kmaher/

@derek it is, uh, definitely non-rigid

RE: https://hachyderm.io/@edfloreshz/116053242210223419

A wild milestone.

@ntnsndr OK, this is going to finally force me to ask the question I've wondered about for 20+ years: why was it called First Monday?

After all these years, it still makes me so sad that the Wikimedia community, that will roast the Foundation at the drop of a hat for doing anything proprietary, does most community chat on *Telegram*. Just gross.

It isn’t just that it is proprietary, though it is that. It’s that we spent a decade fighting to keep Wikipedia safe from the US government, often getting lectured for not doing *enough*, and the community just says “sure we’ll use the platform obviously compromised by Vladimir Putin”. Just embarrassing.

Oh, and I forgot the most damning point: at the Foundation we went to the mat to defend Russian Wikipedians against the Russian state’s harassment. In other news:

“Telegram positioned itself as an anti-surveillance advocate and was to be blocked by the Russian telecommunications regulator. However, the regulator reversed course…noting Durov’s willingness “to counter terrorism and extremism” – which Moscow defines in increasingly broad terms.“

https://www.justsecurity.org/99869/telegrams-security-sham/

@ntnsndr yeah, lots of levels of fascinating in there.

@skinnylatte had a young SF native tell me she was born in the Marina, but “before it was like that”

reader, she graduated high school in the mid-20teens

Most pure expression of “gentrification is what happened after I got here” there could ever be

@scott @skinnylatte it's a very nice place to go for todder's soccer games and to eat at Greens

I am sure there are other things that happened there in the 2020s but I wouldn't know

@glyph I suspect the best argument for facially-neutral licenses is exactly the “messy coalition” part. Easier to get people into the tent if, when the coalition shifts, they still have *something* they can keep using.

I’m not sure that’s a *great* reason but “does it get people to engage and participate” is pretty key.

@whitequark @soph @zkat @joepie91

That's unfortunate, because credible enforcement strategies aren't easy or cheap, which means getting into the game is harder. But if you don't have a public, documented plan, why should anyone author code under your license other than ~vibes~, and why should anyone comply?

@whitequark @soph @zkat @joepie91

@whitequark the *uncertain risk* of GPL lawsuits, combined with low-cost and high functionality, drove a lot of actually pretty good behavior in the early days. No one needed to get sued, that threat just had to be floating out there to get people to the table.

But increased certainty about what the licenses mean, and increased certainty that litigation is very rare, have mostly taken that implied threat off the table. Which, yeah, gets us to where we are now.

@soph @zkat @joepie91

For several years now, I've told groups who want to launch new licenses that day 1 table stakes is "have an enforcement strategy, and talk about it". If you just launch a license and you clearly have no enforcement strategy, might as well not have launched it - neither activist coders nor bad actor users will take it seriously.

@whitequark @soph @zkat @joepie91

@whitequark @legoktm @zkat various patent rights were zapped like this during both WWI and WWII, both times by acts of Congress. There also exist other, more current, exceptions for government use of patents ("compulsory license"), though rare.

As far as I know/can tell, copyrighted works have never previously been considered militarily necessary. Brave new-ish world.

@legoktm @zkat @whitequark licenses basically only stop users with sophisticated compliance departments, and many of the worst users either (1) have no compliance departments (stalkers, cops) or (2) have such sophisticated compliance departments that they will spend a lot of time and money to find loopholes (FAANG).

Which isn't to say licenses are bad (they can be a useful piece of a bigger strategy) but their impact is limited by nature.

We probably need a lot of complicated, sophisticated, highly detailed AI regulation. And any legislation this brute-force will of course need a lot of careful judicial interpretation after it is passed.

But in the meantime, this can be done, can be done swiftly, and addresses a lot of problems.

This whole Rob Pike thing is a good reminder that there is a good, simple, straightforward regulatory proposal out there that would address this and many other LLM problems:

https://crookedtimber.org/2023/05/22/ban-llms-using-first-person-pronouns/

To be clear, this isn’t a pro- or anti-AI take.

The take is “give a human AI to improve ‘browsing’” is “give a fish 16 gears to improve their fixie” territory. Not even wrong.

Figure out what you want to enable people to do, then whether AI is or isn’t part of the solution.

“browse” ain’t it.

🔥 take: Mozilla’s challenge isn’t leadership or AI.

Virtually no one wants to “browse” or to have a “browser”. They want to search, scroll, or use a webapp—if they think about the renderer’s UX something has gone wrong.

So as long as Moz self-identifies as a “browser company”, they’ll lose.