Post #3424130
2026-06-26 14:57 UTC
@alexband@hachyderm.io Nothing in your description has to do with whether the submitter has used an LLM tool or not. You can get slop PRs fully manually typed and great PRs where the contributor has used an LLM.
You just described why the policy should deal in quality, not tool usage.
(LLMs didn't get good at coding until the last half year or so. Previous experiences are not valid - I didn't consider them useful myself until recently when I challenged my opinions)
Replies (2)
-
@mezza@mastodon.pub.solar 2026-06-27 08:32
@troed@swecyb.com @alexband@hachyderm.io stop being so iffy about this. If you want to use slopware, you can just fork the project and vibe code all the changes you want. Just don't expect someone else to put in manual effort when you yourself cannot be bothered.
-
@barubary@infosec.exchange 2026-06-27 10:56
@troed@swecyb.com @alexband@hachyderm.io (LLMs didn't get good at coding until the last half year or so. Previous experiences are not valid - I didn't consider them useful myself until recently when I challenged my opinions) Your whole argument is "I am a good person/developer, and I recently started using LLMs, therefore LLMs are good now".