4 ms·
Emacs also seems to be adapting just fine to the LLM tooling as well.
by uludag 3y ago
Emacs also seems to be adapting just fine to the LLM tooling as well.
- ParetoOptimal 3y agoHere is a very impressive and easy to follow demo of gptel: https://www.youtube.com/watch?v=bsRnh_brggM https://www.youtube.com/watch?v=bsRnh_brggM
- ashton314 3y agoSeriously. Within a month, there were like five different packages that could interface with various LLMs.
- xenodium 3y agoHere’s mine :) https://github.com/xenodium/chatgpt-shell https://github.com/xenodium/chatgpt-shell
- TeMPOraL 3y agoThank you for making it, I did use it quite a bit early on, and likely will again in the near future!
- xenodium 3y agoGlad to hear it! Thanks! Try the experimental compose interface https://xenodium.com/a-chatgpt-shell-compose-ux-experiment https://xenodium.com/a-chatgpt-shell-compose-ux-experiment I found myself using it the majority of the time.
- TeMPOraL 3y agoThough in true Lisp development fashion, they solved the easy and shiny 80% of the problem, leaving rest as an exercise to the reader :). I need to review the current state, but back about half a year ago, when I evaluated them all, I had a lot of "fun" getting them to talk to Azure OpenAI (instead of OpenAI). Or even to allow to override the system prompt. Granted, those were early GPT-4-by-invite-only days, but still. Still, the most important takeaways are that: - Emacs community can and does react quickly to any new development; - Sure, early on the packages may be somewhat half-baked, but it's not that hard to improve/fill in/fix them on your own, with config tweaks and maybe a pinch of simple Emacs Lisp code, even if you barely know anything about programming. That's in contrast to just about any other editor or program, where this is just not possible.