Sure, I'd love to use an open source model locally that's trained on fully ethically-sourced data... once it's on par with ChatGPT or Sonnet.

For now, that "crap" big tech throws at us outperforms the "non-crap".

I'd love to see competition though, K2 is a nice step.