‘Thirsty’ ChatGPT uses four times more water than previously thought - eviltoast
The massive computer clusters powering artificial intelligence consume vast quantities to answer the world’s queries, but how is Big Tech redressing the balance?

Writing a 100-word email using ChatGPT (GPT-4, latest model) consumes 1 x 500ml bottle of water It uses 140Wh of energy, enough for 7 full charges of an iPhone Pro Max

  • areyouevenreal@lemm.eeEnglish
    11·
    27 days ago

    If and until the abilities of AI reach the point where they can compensate tech illiteracy and we no longer need to worry about the exorbitant heat production, it shouldn’t be deployed at scale at all, and even then its use needs to be scrutinised, regulated and that regulation is appropriately enforced (which basically requires significant social and political change, so good luck).

    Why wouldn’t you deploy that kind of AI at scale?

    To be honest I think people keep forgetting that AI strong enough would be smarter than a human, and would probably end up deploying us at scale rather than the other way around. Terminator could one day actually happen. I am not even sure that would be a bad thing given how flawed humans are.

    • lennivelkant@discuss.tchncs.deEnglish
      1·
      27 days ago

      AI strong enough would be smarter than a human

      General AI might be, but the type of “AI” we have right now isn’t general, isn’t smarter, it’s just a really expensive imitation engine that people keep mistaking for actual intelligence.

      And the energy consumption and heat production are really not what our global situation needs right now.

      • areyouevenreal@lemm.eeEnglish
        11·
        26 days ago

        AGI and ASI are what I am referring to. Of course we don’t actually have that right now, I never claimed we did.

        It is hilarious and insulting you trying to “erm actually” me when I literally work in this field doing research on uses of current gen ML/AI models. Go fuck yourself.

        • lennivelkant@discuss.tchncs.deEnglish
          1·
          25 days ago

          AGI and ASI are what I am referring to. Of course we don’t actually have that right now, I never claimed we did.

          I was talking about the currently available technology though, its inefficiency, and the danger of tech illiteracy leading to overreliance on tools that aren’t quite so “smart” yet to warrant that reliance.

          I agree with your sentiment that it may well some day reach that point. If it does and the energy consumption is no longer an active concern, I do see how it could justifiably be deployed at scale.

          But we also agree that “we don’t actually have that right now”, and with what we do have, I don’t think it’s reasonable. I’m happy to debate that point civilly, if you’re interested in that.

          It is hilarious and insulting you trying to “erm actually” me when I literally work in this field doing research on uses of current gen ML/AI models.

          And how would I know that? Everyone on the Internet is an expert, how would I come to assume you’re actually one? Given the misunderstanding outlined above, I assumed you were conflating the (topical) current models with the (hypothetical) future ones.

          Go fuck yourself

          There is no need for such hostility. I meant no insult, I just misunderstood what you were talking about and sought to correct a common misconception. Seeing how the Internet is already full of vitriol, I think we’d all do each other a favour if we tried applying Hanlon’s Razor more often and look for explanations of human error instead of concluding malice.

          I hope you have a wonderful week, and good luck with your ongoing research!