IF I WANTED CLAUDE TO CHANGE MY CODE I WOULD ASK IT MYSELF, I DON'T NEED SOMEONE WHO KNOWS NOTHING ABOUT THE PROJECT TO ASK CLAUDE TO MAKE CHANGES THEY DON'T UNDERSTAND
I sent this to LinkedIn because I like to imagine the rage it induces may be at least proportional to the rage I feel at reading LLM slop
@soller I would check it out on LinkedIn but alas ever time i go there i get bombarded with AI bots in my DM's as well as offers for fake jobs who's descriptions are made by AI to harvest CV's for datamining
@soller I can’t wait for how many non-technical ai bros pretend to be pleasant telling you you’re wrong.
@soller can send them to https://406.fail/
RFC 406i - The Rejection of Artificially Generated Slop (RAGS)

@gfaster @soller thanks I had not come across this response webpage 🀣
@soller this in a nutshell. Why tf would I want your PR when I can ask $LLM myself,
@soller se foes for ANY question. And if I *wanted* an LLM "answer" I would have gotten one.
@soller I wonder if something like Claude could be used to write test programs based on the documentation, both to test the software and to check the documentation. For C programming, I'd put documentation in header files and use that instead of looking at my code. If I had to look at my code to see how to use it, I'd update/clarify the documentation, which then got better over time. It was more work at the start, but saved effort later and made the software easier for others to use.
@soller just thinking of all the karma farming going on across internet coding community sites now driven by people who just copy paste the question into a bullshit generator and copy paste the result back into the thread as if it were real advice... some of them are straight up vibe coding bots to do it for them to save them even that effort. Cultivating an entirely false appearance of helpfulness for social credit has never been easier!
@soller BTW it's what openai do - they don't take prs and ask for clean issues to skip reviewing slop
@soller for πŸ‘ goodness πŸ‘ sake πŸ‘ !
@soller I do see a lot of people complaining about coding LLMs without arguments. But this is a point perfectly made. Thank you!
What a crazy timeline we live in... At my job people are even doing code reviews with LLMs without even trying to understand the code