Project maintainer: "we're struggling to review all this slop-code"
AI bro: "have you tried using a slop generator to review code? I'm very smart"
I am fucking sick of these wank puffins
Project maintainer: "we're struggling to review all this slop-code"
AI bro: "have you tried using a slop generator to review code? I'm very smart"
I am fucking sick of these wank puffins
Code goes in the Claude hole
@Dangerous_beans watched a short presentation from a colleague where he tried to use an LLM to achieve a task that had a 5000 line output. It did ok first pass, oh but a couple things wrong, then he tried to fix those and it just completely stuffed the task. Maybe if I do X and Y, oh no, its still not right. I'll start again, and got a different answer
my conclusion: this tool does not work they way you want it to work.
his conclusion: we need to keep trying different things to see how to make it work