3 ms·

I predict LLMs will reach superhuman level and beat even that model in the next 12 months

• orbital-decay 7d ago
  Starcraft is APM-dependent. Unless the latency will improve greatly in frontier reasoning LLMs (which is unlikely), it will remain a bit like knitting with an excavator.
  • benswerd 7d ago
    I predict latency will improve greatly in the next 12 months to more than 4x speed on current frontier tasks
    • loeg 7d ago
      Yeah but Starcraft needs, like, 10-20x the APM these agents are doing.
      • benswerd 7d ago
        I’m not convinced a lot of it can’t be solved with code mode. Marine staggering for example seems like an ideal code mode task.
        • loeg 7d ago
          Yeah. Some of it may just be "thinking" less rather than faster token generation.
        • jayd16 6d ago
          Do the humans get to use this auto-stagger too?
    • swiftcoder 6d ago
      The best sc2 bots these days play in the ~50,000 APM range (they could mostly go higher as well, but the game client breaks somewhere around 100k APM). I don't see an LLM-based bot getting up to that sort of speed anytime soon. On the other hand, I do think LLM-based bots will quickly outperform the decision-making of many of the hand-coded bots, so maybe they won't need so much APM to be competitive.
  • jamiequint 7d ago
    Should be possible to play this with Jev.
    • bgandrew1 6d ago
      You mean Laya? Jev is not relevant anymore