• phutatorius@lemmy.zip
    link
    fedilink
    English
    arrow-up
    15
    ·
    2 days ago

    Jevons Paradox

    It’s odd (paradoxical, even?) to call it a paradox when that’s what has happened in the medium to long term with nearly every previous period of major technology change.

    I think the problem is that technology isn’t the main driver of how society’s wealth is distributed in a population: that’s a political choice. If technology creates more wealth in an unequal society, those at the top will be the beneficiaries. If it creates a wider awareness of the need for redistribution, then it’ll become a tide that raises all boats.

    The bigger question is whether the LLM bubble is going to have any lasting impact.

    • Wander@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      2
      ·
      2 days ago

      I don’t think so.

      Everything else in thr economy will run and cheap data centres will be used eventually. We use more and more every year the bubble won’t stop thr growth

      • Photonic@lemmy.world
        link
        fedilink
        English
        arrow-up
        8
        ·
        2 days ago

        Not in the long run, but we can expect a period of substantial economic downturn, even more than we are experiencing today. And the economy is highly vulnerable due to concentration of wealth, multiple wars across the globe with serious ramifications for the countries who are not at war and an ever increasing dependency on AI.

        • Wander@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          1
          ·
          22 hours ago

          Economic downturn from oil yes.

          But if chatgpt stopped existing now do you think that was cause a recession? I don’t see it. Its hype but the rest of the economy moves forward.

          The oil through, that’s another factor all together. We are fucked!

          • Photonic@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            ·
            edit-2
            20 hours ago

            It’s not as much about the usefulness of LLM’s (although a lot of companies are more dependent on it every day), but about the not-as-useful as we expected part. Investments in LLMs have gone through the roof. When there are huge investments that do not match the usefulness of a product there is a bubble that bursts. For example tulip bulbs aren’t particularly useful either, but on a much smaller scale the insane investments did cost a bunch of people a lot of money. Another example is the dot com bubble of the late 1990’s.

            So it’s not about ChatGPT ceasing to exist, it’s about overvalued stocks.

            The oil comes on top of that and is what I meant when I said that it would be worse than today, since we are already starting to feel the effects of the oil crisis.

            The oil crisis makes us that much more vulnerable to a market crash and global recession when the bubble bursts.

        • pinball_wizard@lemmy.zip
          link
          fedilink
          English
          arrow-up
          3
          ·
          edit-2
          1 day ago

          Yes!

          And I always add that the financial issue with dependency on AI is that there’s no way to know today what the final capacity will be after the excitement settles down.

          I see plenty of fear of missing out, but less fear of becoming dependant on a suddenly scarce resource. I think the second risk is higher (based on past technology experiences).