-open and enable new folks, and new types of folks to enter and thrive. I think openness and new perspectives are important and healthy for a community.
When you have major community members of significant distros saying that having pronouns in your bio is disqualifying to run for leadership-
I consider Doll to be one of the leading edge experts in agentic coding, so I'll happily take its opinions as authoritative.
I'm not making the case that it's deliberate in general, I'm saying that it's an accepted cultural norm that isn't really noticed by those inside the community. I don't count myself as an insider because, while I am capable and know some things, my background is not in CS.
The degree of accessibility of those varies, for sure.
The initial point was that software has an especially open and sharing culture. I mostly agree with this, and simply tried to raise the point that there are still issues and tendencies that lead to backlash against AI increasing access.
A clarification I can make to my initial statements, now: the technical knowledge might be freely, and often easily accessible, but the social knowledge of how to assemble that knowledge well, especially with others, is not at parity.
I come from a very weird background that makes me very attuned to specific issues in social spaces, and specific knowledge gaps. I don't this gives me privileged insight, just a different perspective that's hard to see if you know the basics already.
-and they aren't summarily repudiated from the community, I think that's an indicator that there's absolutely still work to do. The lovely thing about the software community is that things like that are often more known, and sometimes the tools exist to do things about it.
It means literally anything more quantized than full, unquantized GLM 5.2. So yes, those would be mutilated quants.
Software development is not a fully open, purely meritocratic field. I think it is closer to one than many others, and has some of that ethos structurally and culturally baked in.
To believe that it is a solved issue or doesn't have hidden hurdles is to stop the constant, necessary work to keep it-
I hang around the outskirts of some circles where the basic line is that AI is categorically evil, and against the CoC/contribution guidelines. That recognizing the capability that LLMs have developed means that Don Knuth has totally lost it and can no longer be trusted about anything new.
