• Komodo Rodeo@lemmy.world
      link
      fedilink
      arrow-up
      50
      arrow-down
      1
      ·
      3 days ago

      You’re going to be happy about its application in biochemisty, genetics and astronomy. The volumes of data needing analysis are staggering, they’re just the perfect tasks for a robot brain.

      • dumples@midwest.social
        link
        fedilink
        English
        arrow-up
        22
        arrow-down
        5
        ·
        3 days ago

        Different than LLM. It’s other machine learning that are just using the AI label now to join the hype train. Not you same as what openAI and other AI companies are selling

        • saimen@feddit.org
          link
          fedilink
          arrow-up
          17
          ·
          3 days ago

          But… isn’t it the other way around? LLMs are just one type of AI and they used the overall term AI to hype their products because it sounds better.

          • dumples@midwest.social
            link
            fedilink
            English
            arrow-up
            6
            arrow-down
            1
            ·
            3 days ago

            Kind of. AI is now the catch all now but that’s because LLM were called generative AI. So the older terms all followed suit to catch the hype train. But most of the problems with AI are LLM specifics.

            • partial_accumen@lemmy.world
              link
              fedilink
              arrow-up
              13
              ·
              edit-2
              3 days ago

              Kind of. AI is now the catch all now but that’s because LLM were called generative AI. So the older terms all followed suit to catch the hype train. But most of the problems with AI are LLM specifics.

              Generative AI is a specific subset of Deep Learning, which is a subset of Machine Learning, which is a subset of AI. Generative AI includes more than just LLMs. It also includes the image and video AI creations. There are also GenAI than makes music.

              Machine Learning is typically creating versions of things that existed prior, or recreating things from behavior extracted from prior data. Deep Learning is similar but employs neural networks of various types. The distinction between GenAI and Deep Learning is that GenAI creates something that never existed before, essentially original works.

              Your base point in this conversation stands though. The Deep Learning used in science for analyzing patterns or behavior in, say, biology, is very different (and its more useful) than GenAI and the slop it can produce.

              • dumples@midwest.social
                link
                fedilink
                English
                arrow-up
                5
                ·
                2 days ago

                The chart is good and all but I’ve been a data scientist now for almost 10 years. Most of these terms have changed more then once. There are mostly marketing terms. I’ve done deep learning, machine learning, data mining (my favorite term), statistical learning, predictive analysis and now AI while using mostly the same things.

      • skuzz@discuss.tchncs.de
        link
        fedilink
        arrow-up
        8
        arrow-down
        1
        ·
        3 days ago

        ML has existed for a while, it has been in phone camera software for 7+ years now, for example. The way everything keeps getting lumped in this “AI” buzzword is hilarious.

      • Fedizen@lemmy.world
        link
        fedilink
        arrow-up
        2
        ·
        edit-2
        3 days ago

        Its an imperfect solution to an otherwise difficult task. I don’t know if anyone in science should be happy to use it since its results aren’t generally reproducible (a problem that already existed in biochemistry and genetics) but I could see it as being helpful at narrowing down datasets too large for people to handle.

        I think with this epstein stuff too if the code and training data is not open there’s always a possibility AI has been set to filter something (such as Trump’s name).