Mind-reading AI can translate brainwaves into written text: Using only a sensor-filled helmet combined with artificial intelligence, a team of scientists has announced they can turn a personā€™s thouā€¦::A system that records the brainā€™s electrical activity through the scalp can turn thoughts into words with help from a large language model ā€“ but the results are far from perfect

  • Not_mikey@lemmy.world
    link
    fedilink
    English
    arrow-up
    13
    arrow-down
    6
    Ā·
    11 months ago

    Calling llm ā€œbig auto-completeā€ is like calling people ā€œbig bacteriaā€ . Itā€™s true that they act on the same goal, guess the next word for llm and auto-complete; survive and reproduce for people and bacteria, but they are vastly different in scale and complexity.

    Also what would AI be to you if not an llm? Cause Iā€™d say anything that has an SAT score higher than most Americans has some form of intelligence.

    • knightly the Sneptaur
      link
      fedilink
      English
      arrow-up
      6
      arrow-down
      2
      Ā·
      11 months ago

      ā€œBig Bacteriaā€ is a much more accurate descriptor of humans than ā€œArtificial Intelligenceā€ is of large language models.

      This is the same problem we had with IQ testing, what the test measures is not ā€œintelligenceā€, but the ability to retain and process information according to a predefined schema. This requires no intelligence at all, as demonstrated by the fact that a sufficiently large statistical model of human writing patterns can pass the SATs.

      • orgrinrt@lemmy.world
        link
        fedilink
        English
        arrow-up
        5
        arrow-down
        2
        Ā·
        11 months ago

        Iā€™ve always wondered with stances like this, why do you assume that our ā€œintelligenceā€ is much different than that of llms? I mean, as much as we like to feel superior, is there anything that would really prove that our brains donā€™t work in a similar manner behind the curtains? What if we just get input stimuli and our mind is simply the process of figuring out the most likely answers, reactions or whatever, to that?

        I havenā€™t seen anything to that effect, but then again my field of study is vastly different. Iā€™d like to be enlightened certainly!

        • knightly the Sneptaur
          link
          fedilink
          English
          arrow-up
          6
          Ā·
          11 months ago

          LLMs are statistical models of human writing, they only offer the appearance of intelligence in the same fashion as the Chinese Room thought experiment.

          Thereā€™s nothing ā€œintelligentā€ in there, just a very large set of instructions for transforming inputs into outputs.

          A sufficiently advanced model of the human brain can be ā€œintelligentā€ in the same way that humans are, but this would not be ā€œartificialā€ since it would necessarily employ the same ā€œnaturalā€ processes as our brains.

          Until we have a model of ā€œintelligenceā€ itself, anyone claiming to have ā€œAIā€ is just trying to sell you something.

          • orgrinrt@lemmy.world
            link
            fedilink
            English
            arrow-up
            4
            Ā·
            11 months ago

            What I wonder, though, is if it isnā€™t possible to describe human brain, and the nervous system as a whole, as a very large set of instructions for transforming inputs into outputs?

            • knightly the Sneptaur
              link
              fedilink
              English
              arrow-up
              2
              Ā·
              11 months ago

              It could be described that way, but it wouldnā€™t be a very apt metaphor. We arenā€™t simple, stateful input-to-output algorithms, but a confluence of innate tendencies, learned experiences, acquired habits, unconscious motivations, and capable of modifying our own thought processes and behavior on the fly to suit whatever best fits the local context. Our brains encode a model of the world we live in that includes models of ourselves and the other people we interact with, all built in realtime from our observations without conscious effort.

              • orgrinrt@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                Ā·
                11 months ago

                Iā€™m not disputing that our intelligence isnā€™t more sophisticated, but rather that maybe the ā€œintelligenceā€ in llms is not necessarily all that different from ours, just based on different and limited inputs, and trained on a vastly less wide data.

                • knightly the Sneptaur
                  link
                  fedilink
                  English
                  arrow-up
                  1
                  Ā·
                  11 months ago

                  But it is, necessarily.

                  For example, when we make shit up, weā€™re aware that the shit we made up isnā€™t real. LLMs are structurally incapable of recognizing the distinction between facts they regurgitate and the ones they manufacture from whole cloth.

                  You didnā€™t have to consume terabytes of text to build a model for how to form sentences like a human, you did that with a few megabytes of overheard conversation before you were even conscious enough to be aware of it.

                  Thereā€™s no model of intelligence so over-simplified to the point of giving LLMs partial credit that wouldnā€™t also give equivalent credence to the ā€œintelligenceā€ of search engines.

      • Not_mikey@lemmy.world
        link
        fedilink
        English
        arrow-up
        4
        arrow-down
        4
        Ā·
        11 months ago

        This seems like circular reasoning. SAT scores donā€™t measure intelligence because llm can pass it which isnā€™t intelligent.

        Why isnā€™t the llm intelligent?

        Because it can only pass tests that donā€™t measure intelligence.

        You still havenā€™t answered what intelligence is or what an a.i. would be. Without a definition you just fall into the trap of ā€œA.I. is whatever computers cant doā€ which has been going on for a while:

        Computers can do arithmetic but they canā€™t do calculus, that requires true intelligence.

        Ok computers can do calculus, but they canā€™t beat someone in chess, that requires true intelligence.

        Ok computers can beat us in chess, but they canā€™t form coherent sentences and ideas, that requires true intelligence.

        Ok computers can form coherent sentences but ā€¦

        Itā€™s all just moving the goal post to try and preserve some exclusively human/organic claim to intelligence.

        There is one goalpost that has stayed steady, the turing test, which llm seems to have passed, at least for shorter conversation.

        • knightly the Sneptaur
          link
          fedilink
          English
          arrow-up
          5
          Ā·
          edit-2
          11 months ago

          This seems like circular reasoning. SAT scores donā€™t measure intelligence because llm can pass it which isnā€™t intelligent.

          The purpose of the SAT isnā€™t to measure intelligence, it is to rank students on their ability to answer test questions.

          A copy of the answer key could get a perfect score, do you think that means itā€™s ā€œintelligenceā€ is equivalent to a person with perfect SATs?

          Why isnā€™t the llm intelligent?

          For the same reason that the SAT answer key or an instruction manual isnā€™t, the ability to answer questions is not the foundation of intelligence, nor is it exclusive to intelligent entities.

          You still havenā€™t answered what intelligence is or what an a.i. would be.

          Computer scientists, neurologists, and philosophers canā€™t answer that either, or else weā€™d already have the algorithms weā€™d need to build human-equivalent AI.

          Without a definition you just fall into the trap of ā€œA.I. is whatever computers cant doā€ which has been going on for a while:

          Exactly, youā€™re just falling into the Turing Trap instead. Just because a company can convince you that itā€™s program is intelligent doesnā€™t mean it is, or else chatbots from 10 years ago would qualify.

          There is one goalpost that has stayed steady, the turing test, which llm seems to have passed, at least for shorter conversation.

          The Turing Test is just a slightly modified version of a Victorian-era social deduction game. It doesnā€™t measure intelligence, but the ability to mimic a human conversation. Turing himself acknowledged this: https://www.smithsonianmag.com/innovation/turing-test-measures-something-but-not-intelligence-180951702/

          • Not_mikey@lemmy.world
            link
            fedilink
            English
            arrow-up
            2
            arrow-down
            4
            Ā·
            11 months ago

            computer scientists, neurologists, and philosophers canā€™t answer that either, or else weā€™d already have the algorithms weā€™d need to build human equivalent A.I.

            I think your mixing up sentience / consciousness with intelligence. What is consciousness doesnā€™t have a good answer right now and like you said philosophers, computer scientists and neurologist canā€™t come to a clear answer but most think llms arenā€™t conscious.

            Intelligence on the other hand does have more concrete definitions that at least computer scientists use that usually revolve around the ability to solve diverse problems and answer questions outside of the entities original training set / database. Yes doing an SAT test with the answer key isnā€™t intelligent because thatā€™s in your ā€œdatabaseā€ and is just a matter of copying over the answers. LLMs donā€™t do this though, it doesnā€™t do a lookup of past SAT questions itā€™s seen and answer it, it uses some process of ā€œreasoningā€ to do it. If you gave an LLM an SAT question that was not in itā€™s original training set it would probably still answer it correctly.

            That isnā€™t to say that LLMs are the be all and end all of intelligence, there are different types of intelligence corresponding to the set of problems that intelligence is solving. A plant identification A.I. is intelligent for being able to identify various plants in different scenarios but it completely lacks any emotional, conversational intelligence, etc. The same can be said of a botanist who also may be able to identify plants but may lack some artistic intelligence to depict them. Intelligence comes in many forms.

            Different tests can measure different forms of intelligence. The SAT measures a couple like reasoning, rhetoric, scientific etc. The turing test measures conversational intelligence , and the article you showed doesnā€™t seem to show a quote from him saying that it doesnā€™t measure intelligence, but turing would probably agree it doesnā€™t measure some sort of general intelligence, just one facet.

            • nevemsenki@lemmy.world
              link
              fedilink
              English
              arrow-up
              4
              Ā·
              11 months ago

              LLMs donā€™t do this though, it doesnā€™t do a lookup of past SAT questions itā€™s seen and answer it, it uses some process of ā€œreasoningā€ to do it.

              The ā€œreasoningā€ in LLM is literally statistical probability of which word would follow which word. It has no real concept of what it talks about beyond the pre-built relationship matrices between words and language rules. Thatā€™s why LLMs confidently hallucinate obvious bullshit time to time - to them thereā€™s no meaning to either truthful or absolute bonkers text, itā€™s just words that should probably follow each other.

              • Not_mikey@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                arrow-down
                1
                Ā·
                11 months ago

                All inference is just statistical probability. Every answer you give outside of your direct experience is just you infering what might be the answer. Even things we hold as verifiable truth that we havenā€™t experienced is just a guess that the person who told it to us isnā€™t lying or has some sort of proof to there statement.

                Take some piece of knowledge like ā€œBiden won the 2020 electionā€ me and you would probably agree this is the truth, but we canā€™t possibly ā€œknowā€ itā€™s the truth or connect it to some verifiable experience, we never counted every ballot or were at every polling station. We ā€œknowā€ itā€™s the truth because more people, and more respectable people, told us it was and our brain makes a statistical guess that their answer is right based on their weight. Just like an LLM other people will hallucinate or bullshit and come on the other side of that guess and assert the opposite and even make up stuff to go along with that story.

                This in essence is what reasoning is, you weigh the possibilities of either side being correct, and pick the one that has more weight. Thatā€™s why science, an epistemological application of reason, is so heavily reliant on statisticsā€¦

            • knightly the Sneptaur
              link
              fedilink
              English
              arrow-up
              3
              Ā·
              11 months ago

              Yes doing an SAT test with the answer key isnā€™t intelligent because thatā€™s in your ā€œdatabaseā€ and is just a matter of copying over the answers. LLMs donā€™t do this though, it doesnā€™t do a lookup of past SAT questions itā€™s seen and answer it, it uses some process of ā€œreasoningā€ to do it.

              Youā€™ve now reduced the ā€œprocess of reasoningā€ to hitting the autocomplete button until your keyboard spits out an answer from a database of prior conversations. It might be cleverly designed, but generative models are no more intelligent than an answer key or a libraryā€™s card catalog. Any ā€œintelligenceā€ they appear to encode actually comes from the people who did the work to assemble the training database.

              • Not_mikey@lemmy.world
                link
                fedilink
                English
                arrow-up
                1
                Ā·
                11 months ago

                This is not how LLMs work, they are not a database nor do they have access to one. They are a trained neural net with a set of weights on matrices that we donā€™t fully understand. We do know that it canā€™t possibly have all the information from its training set since the training sets (measured in tb or pb) are orders of magnitude bigger than the models (measured in gb). The llm itself is just what it learned from reading all the training data, just like how you donā€™t memorize every passage in a book you read, just core concepts, relationships and lessons. So if I ask you " who was gatsbys love interest?" You donā€™t remember the line and page of the text that says he loves Daisy, your brain just has a strong connection of neurons between Gatsby, Daisy , love, longing etc. that produces the response ā€œDaisyā€. The same is true in an LLM, it doesnā€™t have the whole of the great Gatsby in its model but it too would have a strong connection somewhere between Gatsby, Daisy, love etc. to answer the question.

                What your thinking of are older chatbots like Siri or Google assistant which do have a set of preset responses mixed in with some information from a structured database.

                • knightly the Sneptaur
                  link
                  fedilink
                  English
                  arrow-up
                  1
                  Ā·
                  edit-2
                  11 months ago

                  This is not how LLMs work, they are not a database nor do they have access to one.

                  Please do explain how you think they make LLMs without a database of training examples to build a statistical model from.

                  The llm itself is just what it learned from reading all the training data,

                  I.e. ā€œa model that encodes a databaseā€.

                  They are a trained neural net with a set of weights on matrices that we donā€™t fully understand.

                  I.e., ā€œwe applied a very lossy compression algorithm to this databaseā€.

                  We do know that it canā€™t possibly have all the information from its training set since the training sets (measured in tb or pb) are orders of magnitude bigger than the models (measured in gb).

                  Check out the demoscene sometime, youā€™ll be surprised how much complexity can be generated from a very small set of instructions. Iā€™ve seen entire first person shooter video games less than 100kb in size that algorithmically generate hundreds of megabytes of texture data at runtime. The idea that a mere 1,000x non-lossless compression of text would be impossible is laughable, especially when lossless text compression using neural network techniques achieved a 250x compression ratio years ago.

                  • Not_mikey@lemmy.world
                    link
                    fedilink
                    English
                    arrow-up
                    1
                    Ā·
                    11 months ago

                    If LLMs were just lossy encodings of their database they wouldnā€™t be able to answer any questions outside of there training set. They can though, and quite well as shown by the fact you can give it completely made up information that it canā€™t possibly have ā€œseenā€ and it will go along with it and give plausible answers. That is where itā€™s intelligence lyes and what separates it from older chatbots like Siri that cannot infer and are bound by the database they pull from.

                    How do you explain the hallucinations if the llm is just a complex lookup engine? You canā€™t lookup something youā€™ve never seen.