Elektrine lite

← Feed

@levitte@mastodon.nu

Post #4113714

2026-07-26 14:02 UTC

@rkaj@mastodon.nu What you're saying, essentially, is that folks using an AI agent as assistant do not direct their agent, do not review the output, etc. By definition.

Replies (3)

  • @levitte@mastodon.nu 2026-07-26 14:15

    @rkaj@mastodon.nu And we're not even talking about other ways to get assistance... like having and agent find a bug that had been haunting me for a few weeks.

    Open ##4113713

  • @rkaj@mastodon.nu 2026-07-26 14:27

    @levitte@mastodon.nu No, I'm saying that the direct LLM output is slop. It is entirely possible for people read that slop, be inspired by it and write code that is not slop. You may even call that "reviewing the output" (it it's a very thorough review). I may not think that is a good idea, for numerous reasons, but it is possible for a human to use an LLM in some ways and still write code that is not slop. Note that Codeberg did not ban all use of LLMs, they just restricted it. https://blog.codeberg.org/protecting-our-floss-commons-from-llms.html

    Open ##4113967

  • @ahltorp@mastodon.nu 2026-07-26 14:27

    @levitte@mastodon.nu @rkaj@mastodon.nu I would need to have a scenario explained to me where an LLM is used to generate code and a subsequent review where the person walks away with as much understanding of the code as if the person had written the code themselves.

    Open ##4113968