Rant: LLMs contributing to FOSS
LLMs contributing is an oxymoron.
If LLMs produced good code, and if LLM-using-humans were honest, polite, and respectful, would it *then* be a good idea for Free Software projects to accept their contributions?
I think the answer is no, not under any circumstances, as a matter of principle.
I wonder if this focus on things like (i) the code being poor, (ii) the maintainer burden increasing due to an increase in (sloppy, huge) PRs, (iii) murky licensing issues, might somehow sidestep the heart of the issue.
These LLM companies have hoovered up every line of FOSS code they could get their hands on, everything that FOSS people have worked on for decades, often unpaid, often underappreciated, often as a labour of love, often explicitly released as a service to the common good.
So they come in, they hoover that up, and tell CEOs "you can fire your workers!", while giving workers the carrot and the stick: if you use this you'll be god-like, if you don't use it you'll be unemployed. They tell governments, if you don't use it, your adversaries get an advantage.
It's non-optional, you have to accept us as middlemen (yes, men), it's the future, etc.
And the problem maintainers have is the code is poor?
The problem is that these silicon valley goons & associated billionaires and politicians view human beings as statistics at best, and as expendable detritus at worst.
Computers and networks, for these people, are a means to extract wealth from populations, and that's all.
FOSS projects should say no to this stuff because LLMs represent a massive shift towards *even more aggressively anti-human tech*.
Sigh, ok, rant over, thanks