• 18 Posts
  • 51 Comments
Joined 6 months ago
cake
Cake day: April 20th, 2026

help-circle






  • A bit unrelated, but I wonder if I’m the only one left that is:

    • Using small, open-weight models
    • Asking it to only write small snippets that I can easily go through, check and understand.
    • Copying pasting from good old web frontends instead of integrating it into my editor or whatever else people do with agents nowadays.

    I feel like when we’re talking about AI this sort of usage gets lumped into the same pile as agentic vibe coding and it’s not helpful for discussion. Surely most people wouldn’t find this problematic?





















  • I get the appeal of this, but I feel this is part of a broader problem of relying on AI for problems that already have elegant, simple solutions. Like relying on LLM’s to tell you what was the last google doc that you viewed is just not sustainable and creates bad habits. Soon people will ask what was the last tab I was just in, instead of using a keyboard shortcut and or having proper tab management.

    Imagine using it for your code editor, instead of using the built-in functionality of jump-lists, marks, fuzzy finders and so on… Or switching to an editor that has these things. What file was I just in? Where did I just write this function?

    “When asked, the AI even double checks if the document was shared in a group chat with his coworkers.”

    How about learning to just use search?