pleroma.debian.social

pleroma.debian.social

Debian has voted to not reject use of AI for contributions to the distribution. I've seen people blaming the voting system, political centrism, or 'being tricked' by AI-written proposals to rectify this result. I will instead blame the voters, the Debian Developers themselves, for the outcome: There simply was no majority to reject genAI contributions, and likewise no interest in regulating it. >>

However, in spite how the winning proposal tries to weasel around the issue, not rejecting AI is an endorsement for use in the distribution (as easily can be seen in the uncriticized use of AI to a) write the proposal, b) submit patches to the voting software and c) analyzing the vote results with a script.).

@leah Both can be true at once. Voters make the results, yes, but the zeitgeist that makes voters vote the way they do is culturally enforced top-down (by the voting system, political centrism, and media, including LLM-written text).

Democracy is really difficult to protect, because people are part of a system, and individual voting does not, unfortunately, tell the whole story.

@leah

Yes, very disappointing.

@leah I don't know enough about their voting process to judge this. Particularly, how the presence of misleading alternatives affects the outcome under whatever system they use. (Arrow's Theorem nerd stuff.)

@dalias There were many pro and and many contra options, and the pro options dominated all the other ones, afaics.

@leah Welp, I won't let this change my opinions of the ethics of the Debian community. https://www.jwz.org/blog/2016/04/i-would-like-debian-to-stop-shipping-xscreensaver/

@leah RIP Debian and the stability it was known for

@leah As a debian developer, this is correct. The results of this vote is shameful, even if some devs essentially voted for it as a "don't come to any decisions right now" option.

@jwz @leah Your quote from 2016 reads like it was written today. ”I guess you want Debian to be the kind of operation that uses the work of others while blatantly and explicitly ignoring the wishes of the person who *did the actual creative work*.” https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=819703#168

@leah 🤮

And unfortunately it's a secret vote, so there's no data on who is untrustworthy.

@dalias @leah Ergo the whole project is now untrustworthy. Look for the people who bail after this, like those after systemd adoption.

@leah

ignore all the people who are telling you to literally "eat sh*t and die" because AI may possibly be in the distro

they are so vile

@dalias @leah

That voting system looked weird to me. As someone who manages the inkscape STV voting machinery it was.. er, strange.

It looks like each two options were ranked against each other; though not with any mathematics I've seen before.

Example: https://inkscape.org/*leadership-committee/elections/board-elections-2015/

@leah looking forward to weird bugs and strange software behavior in all of my debian systems in the near future. ah, what benefit this "tech" to humanity is.

@doctormo @dalias @leah @werdahias might know why. Also the first option, which was for the AI ban, needed a 3:1 majority to be accepted. Was that because it was to be part of the social contract?

@leah
Don't they get funding from the Free Software Foundation? They've voted to allow non-free licensed work to become part of it, because of what the LLMs were trained on. Also the non-free labour of exploited workers. Will that foundation care?

@nivex @dalias @leah It's disappointing. But what does it matter what distros do if the kernel itself accepts them, so they can't possibly be genAI code free, unless they fork the kernel?

@neongod @dalias @leah It matters because the line has to be drawn *somewhere* and rather than pick up a piece of chalk, every organization with enough clout to matter is taking a step back and saying "not it".

If the harms cannot be completely avoided, they should be mitigated to the largest extent possible.

@leah

I completely agree: some issues cannot be reduced to any form of neutrality whatsoever.

Your name is Alice, the wolf is at the door – you don’t leave it ajar.

@tris @doctormo @dalias @leah yes, the 3:1 majority would have been needed because of the social contract amendment.
replies
0
announces
0
likes
2

@neongod @nivex @leah Because FFS the goal is not some mystical "purity".

It's to protect users as best we can from junk code that's going to trash their data (see: rsync slop fiasco) and potentially make them dependent on programs that are no longer free software (and over which there will likely be future copyright status disputes) to access their own computers and data.

@leah
> I will instead blame the voters, the Debian Developers themselves, for the outcome: There simply was no majority to reject genAI contributions

As a Developer [0], I testify this is accurate. We know how to express our preferences through our voting system, and “None of the Above” is always available. The majority have, after much deliberation, chosen a spineless refusal to act.

[0]: effectively means: those who put in the work, and thereby have voting franchise in the Project.