I think we are looking on the wrong angle. We are so focused on the AI usage that we completely ignored part that can bite us hard in the back in the future. This is me trying to untangle this point.

All feedback is welcomed.

all 6 comments

sorted by: hot top controversial new old
[–] 17 points 4 days ago*

I've thought since most of this began that it didn't really matter what he said so much as what he did. And what he did (and continues to do) is commit LLM-generated code.

But reading your post, you're right that it's also extremely concerning that he's using right-wing dogwhistles and moving the kernel project in a more authoritarian direction. And that's largely a what he "said" thing.

I hope the kernel does get forked. I'll switch in a heartbeat.

  • source
  • [–] 3 points 4 days ago* (last edited 3 days ago) (4 children)

    He's clearly trying to privatize the project by leveraging* LLM subscription size.

    All the contributors need to plan on a fork ASAP.

  • source
  • hideshow 4 child comments
  • [–] 1 point 2 days ago (1 child)
  • [–] 2 points 1 day ago

    he literally said that he wants contributors with big LLM subscriptions who can use those expensive models. Which is to say: he wants to only play with the rich kids.

    https://thenextweb.com/news/torvalds-ai-assisted-by-tag-utterly-stupid-open-source-summit

    You don't grasp the arms race. The more slop code is added, the fewer the contributors who understand it will be. The "deep learning" inside the model becomes the same outside.

    What remains is that only contributors paying for the frontier models will be able to work in the codebase.

  • source
  • parent