On Wed, Mar 12, 2025 at 11:10 PM Brent Meeker <meekerbr...@gmail.com> wrote:
> *>> Responsive? That's a test for intelligent behavior not consciousness.* > > *> No it's not. If you shine a light in his eye and his pupil contracts > that's a response. If you ask who he voted for and he says, "Trump" that's > a response.* > *Yes and that proves that the Trump voter is more intelligent than a rock. If I kick a rock I can calculate what the rock will do with F=ma and other Newtonian equations using just a pencil and paper, but if I kick a Trump voter things would be slightly more complicated. * *>> You have made an implicit assumption that consciousness is the >> inevitable byproduct of intelligence. * > > > *> No, I've made an argument that consciousness imagination is an evolved > feature of thought that adds to intelligence.* > *I'm not sure how "consciousness imagination" differs from just "imagination", but computers have been imagining what the next move a competitor might make in a game of checkers since the mid-1950s. And don't you think it would be safe to say that computers reached a new level of intelligence about two years ago? * *>> I think that is a very reasonable assumption and all I ask is that you >> use the same assumption when you judge humans when you judge the >> intelligence and consciousness of an AI. * > > > *> I do. * > *Great, so you must believe as I do that, however imperfect, the Turing test is the only way to detect intelligence (maybe rocks are brilliant but shy and uncommunicative) and consciousness. So computers have been conscious for at least the last two years, or at least as conscious as our fellow human beings are. * > *>> and yet in spite of not being conscious Natural >> Selection nevertheless managed to manufacture consciousness at least once >> and probably many billions of times; the only way that could have >> happened is if consciousness is an evolutionary spandrel, a byproduct of >> another trate that Evolution can see. Intelligent behavior for example.* > > > > *Only because you cherry pick **what to count as "intelligent > behavior".* > *Cherry pick? Use any criteria you want to determine what is intelligent and what is not, all I ask is that you play fair and use the SAME criteria for judging both human intelligence and AI intelligence, and you can't be certain about the motivations of either.* > *> You and evolution can see language behavior such as describing one's > thoughts about a mathematical proof. This is generally taken as evidence of > conscious thought, and that thought as essential to the development of the > proof. Not a spandrel.* *No. Evolution produced us to be intelligent enough to survive on the African savanna and not to be good at making mathematical proofs, that's why we're so bad at it. In the 1950s nobody understood what was intellectually easy and what was hard at a fundamental level. We find it easy to figure out how to move our appendages to catch a thrown ball, or to recognize objects from any angle even under strange lighting conditions, but we find it hard to solve partial differential equations or to play a good game of chess. In 1950 everybody figured that was because one class of tasks was fundamentally more difficult than the other, but when we tried to reproduce both chores from square one we learned that catching a baseball was far more difficult than playing a good game of chess. * *There must be machinery in our head (constructed from genes) that makes even the most clumsy among us to be masters of hand eye coordination compared with today's robots (although that situation is changing at an exponentially fast rate), but there is no such dedicated machinery for being good at chess, so we find that hard. In fact I think it is only a slight exaggeration to say that at a fundamental level a janitor has a more intellectually demanding job (requiring more FLOPS) than a professor of mathematics.* * > maybe you think mathematicians never get laid.* *I don't know this as a fact but I suspect that on average janitors at universities have more children than professors of mathematics at those universities. * > *>> you did say that to conclude that something is intelligent "require >> inferences about their internal motives and intents", and nobody would say >> such a silly thing if AI didn't exist and nobody had even proposed that >> such a thing might someday be possible.* > > *> LLMs require prompts as motivation. * > *And I wouldn't have written this email if you hadn't prompted me to do so in your previous email, and you wouldn't have written that if I hadn't prompted you to do so in my previous email. * *> You claim to know they are intelligent by their response. But you > wouldn't say that if the response didn't match the motivation.* > *I understand the motivation of a modern large language model about as well as I understand the motivation of one of my fellow human beings. I'm sure there have been times when you saw somebody do something very strange so you asked "why did you do that?" and the response you received you did not consider satisfactory. Sometimes the only response possible was "because I wanted to" because the person is unable to explain the detailed pattern of neuron firings that caused him to do what he did. Exactly the same thing could be said about a modern AI. * * John K Clark See what's on my new list at Extropolis <https://groups.google.com/g/extropolis>* rdb -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscr...@googlegroups.com. To view this discussion visit https://groups.google.com/d/msgid/everything-list/CAJPayv3FkjghemgCjLgNk%3D3qHD%3DW-E%3DVsY51iuTz84wFZfiUoQ%40mail.gmail.com.
