10 comments

  • thorum 3 minutes ago
    I regretfully suspect this actually is the future of UI. You just need smart local models running fast enough to keep the framerate high, but that’s inevitable in the long term, isn’t it?
  • yjftsjthsd-h 9 minutes ago
    Wait, how did it `sleep 3`? I thought LLMs didn't have a concept of time
  • psandor 31 minutes ago
    Reminds me of this project from Microsoft. A Quake II “simulation” model where the whole application was just a model.

    https://www.microsoft.com/en-us/research/articles/whamm-real...

  • __mp 27 minutes ago
    It would be interesting to expose this on the internet. I wonder what the scrapers etc… would do with it.
  • oilkillsbirds 38 minutes ago
    Very cool. Fun project. I feel like there’s a wide field to be explored once you understand fine tuning good existing models
  • obilgic 44 minutes ago
    There was a whole fake gui OS with similar concept right
    • prodigycorp 22 minutes ago
      athropic had an experiment that did exactly this. we need more experimentation in this space.
  • prodigycorp 23 minutes ago
    Wow! This this reminds me of when we tried doing this with gpt 3 and.. it worked!

    The tmux thing had me confused for a moment until i realized it was emulating tmux. Super fun. Is the model window width aware?

  • simedw 3 minutes ago
    [flagged]
  • htoz1 40 minutes ago
    Where is the point of this project if i have terminal ?
    • YmFzZTY0ZGFkZHk 20 minutes ago
      Sometimes it's just fun to do something absurd because you can. This is one of the reasons software is exciting.
    • mehmetoguzderin 24 minutes ago
      Terminal world models have many applications, but empirical observation of successful pretraining or fine-tuning can be significant for encapsulating from rather expensive Bash traces, such as ones obtained from systems with scarce accelerators, etc.
    • good-idea 18 minutes ago
      because curiosity is the most fun of virtues!
    • dotancohen 37 minutes ago
      Because it was there