Boosted by cwebber@social.coop ("Christine Lemmer-Webber"):
gray17 ("felix (grayscale) 🐺") wrote:
@cwebber this sentence near the end struck me as funny
> and I see codebases turning more and more into obscure and confusing organisms that can only be diagnosed by more machines.
sure, your bullshit LLM can bullshit an explanation of its bullshit code, and it will apologize profusely if you tell it "no you're wrong"
Clearly the future of coding is repeatedly telling the LLM "you're wrong" until it reaches a stable "no I'm right and here's why" state