The new #LLM debate in #Debian has made clear to me that we loose the incentives to solve

a) our technical problems using deterministic and less energy-wasting classical methods (Lintian)
b) our social problems by talking to each other (teaching upstreams about SPDX and other machine-readable methods to declare copyright)

Of course this is only one facet to the LLM problem (there are larger issues) but I joined Debian because I want to do the right thing and not "paper over" real problems.

@josch : wait, there’s a new #LLM debate in #debian ?

Where is it happening?

@ploum Sarcasm or real question?

@josch : real question. I thought it was closed after the vote. Didn’t see anything on planet Debian.

@josch I don't think it's true. But I think that after the LLM GR, a lot of people just got quieter... myself included

@tachi I think that's not unexpected and I think is normal after every controversial GR. Those who are not happy with the GR have two options a) stay in Debian and deal with doing things in a way that the don't like much (and then, why be loud and continue voicing dissatisfaction, you already know that your option "lost") or b) leave (and then there is no reason to say much either)

Whichever way such a GR ends just decides who it will be become quiet or leave to spend their freetime elsewhere.

@josch I would think that the number of sponsored seats and paid contributors greatly impacted any sort of democracy in play. In order to move a crowd you just need to move 3%. I think a community that is on the wire of a political debate can be saved but don't try to convince everyone. Try and find that 3% that can move it in the right direction. Even machine cost is being weaponized against us. Time to dig in, not lose hope.

@vramvoolenaar Even if Debian ultimately drops into the slop bucket, FOSS is more than just Debian and ultimately, I believe in the ideas of sharing our code and building it together. Doing so does not require Debian in the loop. And I find myself lucky that I made many strong friendships through Debian and these connections are immune to whatever result a Debian GR has, fortunately. :)

@werdahias @josch I would doubt it’s even any good for the medical scans

https://kolektiva.social/@LevZadov/117429643461966639

@werdahias @josch though for breast cancer detection we tend to use simple image recognition models, not LLMs.