@Harblinger Yes but I thought it would be more "fun" to try that on a standard Debian install... Probably a very bad idea

@Caleb James DeLisle @FourOh-LLC @Harblinger I also didn't used it much for programming so far but I progressively use it instead of search engines for everything technical. The time gained over troubleshooting through increasingly worse search results is just phenomenal.

I also learned a ton while using it. It's not magical of course (currently learning to use Pyinfra and half of the suggestions are just "hallucinations"), but it's very accurate most of the time

@Caleb James DeLisle @Harblinger The fact that he is way less sycophantic out of the box than others LLM makes it more enjoyable to work with

Yesterday we tested its creative ability by asking it to generate some over the top polar/film noir story about an alcoholic detective and we really had a good laugh, it nailed the style and tone perfectly

@Harblinger Claude is quite the jokester

@Harblinger @‎ do you think diving in and starting messing with diy eurorack modules is a viable way to learn electronics?

@:blank: @nprofile1q... @Blurry Moon Ashley is completely unhinged but she rocks, her instance is down right now but her vintage gear videos are at https://videos.icum.to/c/toploadedduke/videos

Sadly she removed all of her funny ones

@Harblinger let's go for a Tin Man discography marathon

@Harblinger I much rather use Claude specifically because its answers are terser and strictly relate to what I asked compared to chatGPT profusely verbose output even when instructed to tune that down

@lain @:blank: @sanek @:VD15_0::VD15_1::VD15_2::VD15_3::VD15_4::VD15_5::VD15_6::VD15_7: looked at the source code of Honk last month and giggled at how idiosyncratic the naming choice was

@lain @Harblinger So far I had better success by using Claude or Mistral as a "consultant" for when I have doubts about how I should implement something or how to rewrite part of the program I knew to be shoddy. I learned a lot by doing that. I set up gpt.el in Emacs for code and tend to use the web interfaces for questions that don't need a lot of context

The code produced by Mistral Vibe was getting a bit sloppy toward the end, like if it started missing some of the context along the way. I did nothing special regarding the feedback loop so that's most likely the issue

I did not try anything by OpenAI or Google yet, surely I must miss out but I slowly catch up with the new world