Artur ManuelMar 13

A systemd issue was posted at least a bit ago: https://github.com/systemd/systemd/issues/41085

This proposes getting rid of LLM-generated code from the project. While I have my qualms with how the issue is presented (too little info about how it actually affects the users of systemd IMO), it is a pretty good issue other than that. Feel free to repost to spread this out more.

#tech #technology #linux #systemd #ai

Disallow usage of generative AI to write code · Issue #41085 · systemd/systemd

Component No response Is your feature request related to a problem? Please describe Generative AI is actively killing people, driving up costs, and plagiarizing work from many open source developer...

GitHub
Show threaddonnyMar 13
@amadaluzia While I hate slopmongers with a burning passion, I've found LLMs to be quite nice at quickly solving some problems. A total ban sounds more like an ideological request than anything that has to do with writing good code.
Show threadArtur Manuel
@donny I've used it for problem solving previously as well, but I've not been so lucky. It was a Rosetta Code project, then being written in Zig, and after developing surprisingly slow solutions, I decided to try Claude to create it. Sure, it improved the speed, but slightly. I decided to entirely reshape the solution which ended up improving execution by 700ms. I don't think it's good enough for my use.
Mar 13 at 11:49amWeb
Show threadArtur ManuelMar 13
@donny To imagine this writing code in systemd is frankly absolutely risky, at least IMO. Sure, it may have caught issues as per what Poeterring said, but the long game seems like a really good risk. I wouldn't see it as "ideological".