Replies (3)

• @thufie@social.pixie.town 2026-08-08 14:24

  @algernon@come-from.mad-scientist.club The scoring procedures behind these metrics really tell you a lot about the actual coding practices of "AI" guys when they aren't outsourcing all their thoughts and keypresses... You'd need a movie projector to find a more direct case of projecting. So of course, surprising no actual software engineers or coders, it does not reflect very well on them.

  Open ##5175747

• @algernon@come-from.mad-scientist.club 2026-08-08 14:24

  Okay, okay, just one more toot... I managed to coerce the tool into giving me an overall "which tool is responsible for this directory, at what percentage certainty?", and the iocaine-powder crate is allegedly 49% Claude. (Reality: it's 80% human, and 10% emacs, because I use snippets for the copyright headers, and 10% clippy & cargo fmt, because we're best buddies, and they yell at me when my code is terrible.)

  Open ##5175748

• @gnpar@mastodon.social 2026-08-08 19:46

  @algernon@come-from.mad-scientist.club and it's so easy to check, it's not like non llm code is so hard to find. But then again, why would a slopfondler ever verify anything? "The AI said it, it must be so".

  Open ##5175755