• 0 Posts
  • 16 Comments
Joined 1 year ago
cake
Cake day: June 18th, 2023

help-circle









  • kautau@lemmy.worldtoLemmy Shitpost@lemmy.worldAI Summary
    link
    fedilink
    arrow-up
    0
    ·
    edit-2
    23 days ago

    I mean also AI summarizing conversations is a useful feature as long as you view it as what it is. Potentially one of the few examples where an LLM is actually useful. My family has an ongoing group chat that is constantly just dumb shit; some days just 100 messages ranting about something or other, but sometimes there’s something worth seeing like events and the like. Trusting AI to glean all the information properly is dumb. Being able to use it to quickly grep what’s happening before doing your own reading isn’t.





  • As long as “cleaner than you found it” also includes “better documented.” I’ve worked with people who think that “the code should speak for itself” to the point that they will make biased decisions with no explanation or documentation and then if you ask them about it after their response is “look at the PR for how that decision was made.” I’m not going to git blame and find your PR to find an outcome from an argument between two people that after scrolling just says “sometimes the API returns a JSON string here instead of nested JSON so we have this conditional” when that could be a comment