• 0 posts
  • 21 comments
Joined 2 months ago
Cake day: April 27th, 2026
  • There’s some fine nuance to finding that balance, though. If the word I want to use is less well-known, but better fits what I’m trying to say (or is significantly shorter than “going the long way around”), then I’m more likely to use it. I think most everybody should strive to communicate as clearly as possible, and that may entail learning the occasional new word. Intentionally “talking down” can be insulting, as well as a step towards Idiocracy if we all start doing it together.

    I get that it’s appropriate in certain situations, or for certain audiences, but I try to stretch my brain to keep it learning (and therefore more sharp), and I hope most others strive to better their communication skills as well. IMHO it’s necessary both to succeed, and to evolve.

  • Oh, okay - on your absolutely minimalist say-so the world will stop using LLMs.

    🙄

    If you’re going to make an assertion, at least try to back it up with an explanation or example if you want to be taken seriously. Otherwise you’re just wasting everybody’s time - including yours (and potentially diminishing your reputation).

    In the meantime, I’ll just point out there’s a reason I said to validate what it’s telling you.

  • LLMs (“AI” is quite the misnomer, IMHO) can be useful when its source material is well vetted, or at least made up of mostly reliable input. In the rush to dominate the market, “AI” purveyors have been far less judicious in what they provided as input to their LLMs, and in the end shot themselves in the foot by doing so. Validate what it’s telling you against actual sources, and it can indeed be helpful.

    ETA: Also check for information it leaves out as you’re doing that validation - especially when it involves server administration, because missing details can be quite fatal to your server security.