Tech bosses think nuclear fusion is the solution to AI’s energy demands – here’s what they’re missing
theconversation.com
Nuclear fusion can’t produce net energy, so is it really a solution to AI’s growing energy demands?

nuclear power produces long-lived radioactive waste, which needs to be stored securely. Nuclear fuels, such as the element uranium (which needs to be mined), are finite, so the technology is not considered renewable. Renewable sources of energy, such as solar and wind power suffer from “intermittency”, meaning they do not consistently produce energy at all hours of the day.

fusion technologies have yet to produce sustained net energy output (more energy than is put in to run the reactor), let alone produce energy at the scale required to meet the growing demands of AI. Fusion will require many more technological developments before it can fulfil its promise of delivering power to the grid.

  • gdog05@lemmy.worldEnglish
    106·
    27 days ago

    Maybe AI can help us break the fusion hurdles. Oh. It’s still telling people to eat rocks, just used to create waifu porn and as a mass spy application? Nothing else, really? Well shit.

    • hark@lemmy.worldEnglish
      5·
      27 days ago

      I know you’re being reflexively downvoted by who hate everything AI, but this is the sort of thing AI should be most useful for, which is finding patterns within large problem spaces with many variables.

        • hark@lemmy.worldEnglish
          2·
          26 days ago

          Wow, I completely misread that post. Now I’m wondering if I replied to the wrong message or if my brain just completely crapped out.

    • barsoap@lemm.eeEnglish
      3·
      27 days ago

      Probably. Proxima fusion is using simulation-driven engineering to pave their way through the design space, no matter how you approach it it’s gotta involve dimension reduction in some way and that’s ML. They speak of AI but well it’s a press piece.

      LLMs or diffusion models? Nah, don’t think so. This is actual engineers throwing statistics at a particular problem to identify what prototypes they should build, not techbros throwing shit at the wall.

    • rottingleaf@lemmy.worldEnglish
      21·
      27 days ago

      It’s even bad at porn. Very limited means of describing the process, forgetting that there are no bed sheets in a park, same repeating metaphors. Boring.

    • Eheran@lemmy.worldEnglish
      78·
      27 days ago

      Nothing else? Please do not speak for other people if you can not grasp what others do with this tool.