• trepli@lemmy.ml
      link
      fedilink
      arrow-up
      2
      arrow-down
      1
      ·
      2 days ago

      https://whynoai.info/

      Address one problem and many more are still there waiting to be solved. Pollution is not the only one, if you still wanna state that it’s over blown

      • ReCursing@feddit.uk
        link
        fedilink
        English
        arrow-up
        1
        ·
        2 days ago

        The only one of that page that’s a “maybe” is “it enriches bad people”, and there are more open alternatives anyway. The rest are either flat out wrong or are humans being bad humans, not the fault of ai. Putting it all on one web site does not prove anything

    • skulblaka@sh.itjust.works
      link
      fedilink
      arrow-up
      14
      arrow-down
      5
      ·
      edit-2
      4 days ago

      1

      That’s false. Here’s a Forbes article about it. And here’s the first paragraph of it to make sure you actually read.

      Artificial intelligence does not just run on algorithms, but on electricity, water, and carbon. OpenAI’s training GPT-3 alone consumed 1,287 megawatt-hours of electricity and required more than 700,000 liters of freshwater, according to a University of California analysis. That’s enough power for 120 U.S. homes for a year and enough water to fill two-thirds of an Olympic pool. Additionally, this training generated 552 tons of CO2. When scaled to today’s GPT-4 and GPT-5, training costs soar into billions of liters and terawatt-hours, reshaping local utilities and global emissions.

      It should be noted that good old fashioned capitalist corruption is what brought you all the big AI models in the first place.

      2

      Sure, they do now. Where do you think those models got their original training data from? What percentage of AI users do you think are using these open source, open weight models?

      In addition, here’s a little snippet right from the front page of OpenClaw’s own website:

      Open-weight means the model weights are downloadable. It does not necessarily mean the training data, code or full development process is open, and the license may still restrict some uses.

      Prime example of stupid fucking bullshit uninformed response here. Maybe try shutting the fuck up about shit you know nothing about? No wonder people downvote you, if you’re this much of an asshole all the time.