New blog post, Open Source and AI: https://kusma.xyz/blog/2026/03/26/open-source-and-ai/ #openSource #AI

[kusma@floss.social]

New blog post, Open Source and AI: https://kusma.xyz/blog/2026/03/26/open-source-and-ai/ #openSource #AI

[unkx@icosahedron.website]

@kusma very well put!

[valpackett@social.treehouse.systems]

@kusma a very eloquent summary of everything. a bot could never!!

[khrys@mamot.fr]

@kusma Thanks for that

[mathieui@piaille.fr]

@kusma Thanks, this is a very good post.

[jani@floss.social]

@kusma

The "Keeping the brain active" part reminds me of something I posted way back:

Jani Nikula (@jani@floss.social)

I'll never stop programming. I'll keep coding as long as I'm able to. It's the ultimate brain teaser. I don't care for sudoku and the like; I'd rather write a sudoku solver (again, in new languages) than solve sudokus manually. And, obviously, telling a language model to do that for me has no appeal. It's like asking a person who loves sudokus if they'd like to use a sudoku solver instead. #CoderForever

FLOSS.social (floss.social)

[pinchartl@floss.social]

@kusma A moral aspect you haven't mentioned is that the companies behind LLMs are happy to ignore how their products are heavily used to manipulate public opinion and undermine democracy. Profits being favoured over democratic values isn't new, but I think we're taking it to a new extreme here.