Felicitas Pojtinger 🌅One more anti-LLM-assisted contribution policy based on data or experiences from early 2025 and I will start writing angry letters to the editor /hj
Gosh I wonder why my experience with these things seems to be so utterly different from the like 10 remaining people still pushing this narrative. Is it because they can't accept they were wrong? Or did they just simply never learn how to use it/haven't tried it in 6 months? I just don't understand. Maybe it's just a different ethical framework/schism at this point, who knows.
So so much of this just reminds me of the anti-Kubernetes, anti-IDEs/LSPs, anti-Electron, anti-systemd, anti-Wayland stuff from years ago. I'm so tired of this debate. Not everything is the same as the cryptocurrency/web3 desaster all over again just because some people are involved in both
1peter10@pojntfx I agree that the usefulness and accuracy has improved (dramatically even), but there are other issues beyond functionality.
The ethics of how the LLMs are created remain as problematic (from copyright infringement to poor human labor conditions), the energy use in a planetary crisis remains an issue, and - although that depends on the use-case, and mostly impacts LLM code generation and IMHO less review - the copyright and thus licensing issues are also unsolved.