Daniel Dennett, Lawrence Krauss and Massimo Pigliucci discuss The Limits Of Science

44

 

00:00 Introduction

07:07 Limits of Science

19:40 God & the Supernatural

31:20 Science & Morality

50:11 Something out of Nothing

1:03:42 The Value of Philosophy

1:20:59 Cognitive Limits

1:35:43 Questions:

1:35:56 Science & Politics

1:43:33 The Status of Economics

1:48:17 Does Consciousness Exist?



1:55:00 Credits



English and Dutch subtitles coming soon!



Website: http://www.hetdenkgelag.be

Written By: het denkgelag
continue to source article at youtu.be

44 COMMENTS

  1. Just started watching this. I really like what Krauss starts saying at around minute 11 and then at 12:15 he makes a critical point. Ultimately if something can be called knowledge it’s science. Science doesn’t just apply to understanding the material world, the scientific method can apply to being a scholar of the bible or being a political scientist, it’s the tools: peer review, testable hypotheses, consistency with other areas of science, those are what matter.

  2. This is a superb discussion, long, but definitely worth it. Full marks to the young host who let the eminent guests speak at length, without butting in, and gently introduced a very interesting set of questions and topics.

  3. At around minute 23 Massimo Pigliucci mentions a science fiction novel in the context of the question what would constitute evidence for God based on intelligent design. Another excellent little known novel on that topic is Eye In the Sky by Phillip K. Dick. In the novel a group of people taking a tour of a particle accelerator get warped into their own consciousness. They jump from world to world and each world is based on the beliefs and bias of one member of the group. The world that is generated by the Christian is really interesting. There is no more science or scientific laws governing the universe so arbitrary things can happen for no reason but the will of God. To get your car started in the morning you don’t use a key in the ignition switch you say a little prayer. It’s really a fascinating glimpse at what an absurd crazy place the world would be if theism were actually true.

  4. At around minute 54 Dennett talks about the Turing test. I was rather disappointed that he doesn’t clarify that the popular conception of the Turing test — the one that he in fact uses — is not what Turing actually wrote. The REAL Turing test was what Turing called The Imitation Game, to see if a computer could not just converse using natural language but to see if it could deceive as well as a human. He first conceives of a test between A and B where A is a man and B is a woman (the physical nature of all the participants is of course obscured and they are communicating via a teletype machine) and the goal of A is to convince the evaluator that he is actually a woman. Then he imagines replacing A with a computer and wonders could the computer do as good a job at deception as the human male.

    I also disagree with Dennett’s evaluation that the Turing test was a miserable failure. I see his point that a lot of people (John Searle and countless other philosophers, many of whom have battled Dennett) didn’t get it but that wasn’t a fault of Turing, if anything it highlighted how vapid and wrong headed the opposition was that they couldn’t allow themselves to focus on questions that could actually be answered and had to concentrate all their effort not on answering questions but obfuscating the work of others.

  5. Very good discussion, nice and long. There were at least a half dozen side trips into subjects that could have been entire seminars; it will be worth sifting through for future essay subjects. I wish the consciousness part at the end had come sooner so it could have been explored in greater depth, but what there was did have some content. I doubt consciousness is a specific thing (without behavior), it is just that we can observe more or less conscious behavior in an organism in a given context. You exhibited no conscious behavior when you were something like a tiny 16 cell blob (although complex organizational behavior was going on), but did when you were just a few months from birth. The richness of that conscious behavior continued to develop for many years thereafter. Massimo didn’t need to speculate about rocks; once upon a time the entire Universe was very hot plasma, and no conscious behavior was possible because no patterns of matter could hold together long enough to do anything past bounce around. But then later, here we are.

  6. These men will feel very silly if they are still alive 20 years from now for arguing that man is capable of understanding the answer to any question he can ask. I predict artificial intelligence will discover things that boggle the minds of any humans trying to understand their discoveries. Even a mathematical formula that will not fit on a page will block nearly everyone. We assume the universe is mathematically simple because we first went after the simplest low-hanging fruit. Then there are surely questions too complex for humans to ask.

    If a result is too “hairy” with crucial detail/exceptions, humans can’t grasp it. It needs to be bare.

    • In reply to #7 by Roedy:

      These men will feel very silly if they are still alive 20 years from now for arguing that man is capable of understanding the answer to any question he can ask. I predict artificial intelligence will discover things that boggle the minds of any humans trying to understand their discoveries

      If what you said about AI is true it wouldn’t contradict anything that was said by them. In fact it would support what they said, that humans will continue to find methods and tools to work around what may seem like insurmountable road blocks to the advancement of knowledge.

      I do agree though that they missed an important point in that discussion and Dennett misrepresented the point that Chomsky and other evo-psych people make. The point is that our fundamental concepts such as logic and math are in some sense hard coded into our DNA. Of course they are highly flexible and can be used in all sorts of ways that have no relevance to the function that the original adaptation was designed for but the point is that at some point — or in fact perhaps we will never even be able to be aware of it — our understanding of the universe may be limited by our mammalian brains. But Dennett presented that as if it was some argument for mysticism or not doing further research when it absolutely is not. It’s just an interesting observation about the possible limits of human knowledge.

  7. In reply to #10 by Quine:

    When? I must have missed that. Do you have the time code?

    He starts at 1:24. It’s odd because I loved the little speech Krauss gave right before that on the possible limitations on what we might be able to know in physics. I think you could make the same argument about cognitive science. Not just that there may be limits on what we can even observe in the physical world but that the mechanisms we have to observe the world may also have inherent limitations.

    For example, what if there are 11 dimensions but for some reason viewing more than three has little value for survival so our nervous system just can’t process it? Or what if the common sense notion of cause and effect that a mammal requires to survive really doesn’t make much sense in a unified theory of physics and we have to understand a totally new way of thinking about logic and math that doesn’t depend on causality? I think those are the kind of things that Chomsky has in mind and I don’t understand why Dennett dismisses them.

    • In reply to #11 by Red Dog:

      In reply to #10 by Quine:

      When? I must have missed that. Do you have the time code?

      He starts at 1:24. It’s odd because I loved the little speech Krauss gave right before that on the possible limitations on what we might be able to know in physics.

      I don’t know why he went that way. Maybe he is referring to specific disagreements that I don’t know about, but I can construct plenty of cases, just as you, that would not fall into things we would expect to be able to handle with the structure of the brains that we have. We already have hundreds of people walking around with very different brains than the 99.99999% of the population, whom we would expect to be able to answer questions that are out of reach for the rest of us. Also, Dennett seems to be restricting to symbolic solutions. There will be problems where you can grok an answer that you can’t put in symbols beyond, “well, it’s sorta like …”

  8. In reply to #4 by Red Dog:

    At around minute 23 Massimo Pigliucci mentions a science fiction novel in the context of the question what would constitute evidence for God based on intelligent design. Another excellent little known novel on that topic is Eye In the Sky by Phillip K. Dick. In the novel a group of people taking a…

    I was a bit surprised Prof. Pigliucci mentioned the sci-fi example without qualifying that that in itself is a product of our distinctly human way of knowing considering he seems to be very careful about our human baggage influencing other modes of experience.

    Mike

    • In reply to #14 by McCourt:

      Lawrence should watch his videos, with a mind to counting how often he interrupts his fellow panellists.

      When I watch discussions like this involving Krauss as a panelist, I can visualise him as a kid…full of nervous energy and his brain working at ten times the speed of everyone else.

      I enjoyed the discussion and was happy to put in the time needed to watch.

      • In reply to #16 by Nitya:

        In reply to #14 by McCourt:

        Lawrence should watch his videos, with a mind to counting how often he interrupts his fellow panellists.

        When I watch discussions like this involving Krauss as a panelist, I can visualise him as a kid…full of nervous energy and his brain working at ten times the speed…

        Yeah, that’s great and all, but he misses so much because of his inability to just listen. He’s already jumping onto what he wants to say, before they’re finished talking. It’s an endemic problem with him.

  9. These guys are surprisingly lost on the subjects of morality and aesthetics… it gets a bit sophomoric and embarrassing, frankly, at moments. This Massimo character uses Sam Harris as a “whipping boy” only in his dreams. Harris has a lot figured out that Massimo just fumbles at.

    • In reply to #19 by McCourt:

      These guys are surprisingly lost on the subjects of morality and aesthetics… it gets a bit sophomoric and embarrassing, frankly, at moments. This Massimo character uses Sam Harris as a “whipping boy” only in his dreams. Harris has a lot figured out that Massimo just fumbles at.

      Was there any specific thing someone said you are referring to? Who exactly said it and when?

      • In reply to #20 by Red Dog:

        Was there any specific thing someone said you are referring to? Who exactly said it and when?

        It’s happening at about 48:10

        Did Massimo even read The Moral Landscape? Of course screaming is empirical evidence!

      • In reply to #20 by Red Dog:

        In reply to #19 by McCourt:

        These guys are surprisingly lost on the subjects of morality and aesthetics… it gets a bit sophomoric and embarrassing, frankly, at moments. This Massimo character uses Sam Harris as a “whipping boy” only in his dreams. Harris has a lot figured out that Massimo just fu…

        Just took a lucky stab at the video again, jumped to 37:00-ish, where Dennet takes a relativistic approach to the question of determining “better” and “worse” regarding morality and aesthetics (he uses chess and card games as his example), and Krauss goes on about what he considers “pretty” (aesthetics, again) in relativistic terms, too.

        If Harris were there, he could have pointed to his “Moral Landscape” which has many objective peaks and valleys: the key word being ‘objective’. If Steven Pinker were there, he could point out that humans are not “blank slates” who determine for themselves “better” or “worse”, but simply respond, given their human architecture, to the stimuli of their surroundings as humans will.

        Now, that would be a good discussion: Harris and Pinker on Morality and Aesthetics… make this happen, Internet!

        • In reply to #35 by McCourt:

          In reply to #20 by Red Dog:

          In reply to #19 by McCourt:

          These guys are surprisingly lost on the subjects of morality and aesthetics… it gets a bit sophomoric and embarrassing, frankly, at moments. This Massimo character uses Sam Harris as a “whipping boy” only in his dreams. Harris has a lot fig…

          I think you are referring to Dennett’s discussion of Normative systems of thought? I agree that was lame. I think he was proposing that you could have philosophical or theological models that aren’t scientific but still have their own internal logic and can be said to be knowledge although not science. A classic example would be Freudianism. Some die hard Freudians would claim that while it’s not science it still has it’s own logic and “meaning”. Or other examples would be Marxist analysis or (shudder) Postmodern theories.

          Actually, I was kind of surprised that Dennett would be supporting such things so I may not have followed what he meant at all. If he didn’t mean that I have no idea what he meant. If he was talking about such “alternatives” to science I think he is wrong.

          If you are trying to answer the question is bridge or poker a “better” game in some abstract sense of gaming you just have a poorly defined question and you can’t get a meaningful answer. If you want to talk about the kinds of skills required, the relative complexity in a game theoretic sense, etc. then you are just back to doing a form of science. Not very conventional science but IMO and Krauss I’m sure would agree it still really is science.

    • In reply to #21 by Peter Grant:

      15:38 Lawrence Krauss: “Mathematics is a language, it isn’t knowledge”

      Thank you!

      I think this is a matter of definition… Using only pure math you can solve real problems arising in economics, in fluid dynamics, in biochemistry etc. Also, it is necessary if you want to have a deep understanding of how things work in physics. Maybe I am a little bit biased (as a mathematician myself !) but I just wanted to make the point that is not just a language, and is up to you whether you call it knowledge or not.

  10. 31:20 Science & Morality

    Lawrence is off to a nice start with consequentialism, glad he’s started listening to Sam Harris.

    Massimo proceeds to completely confuse the issue.

    34:50 Lawrence Krauss “Well, I don’t know how you can say that…”

    LMFAO!

    The point the philosophers seem to be missing is that morality is not fully rational, at least not from an individual subjective perspective. That is why we need science to step out of our own subjectively and see subjectively objectively.

    45:38 Daniel Dennett, “No there aren’t other ways of knowing, but there are other ways of doing things”

    Fair enough, but science is better at them.

  11. @Daniel Dennett

    Yes I am familiar with the Turing problem and the hard problem. The Turing problem is real. I’m not even convinced that anyone else is conscious in the hard sense, but I’m willing to give them the benefit of the doubt.

    • In reply to #26 by Peter Grant:

      @Daniel Dennett

      Yes I am familiar with the Turing problem and the hard problem. The Turing problem is real. I’m not even convinced that anyone else is conscious in the hard sense, but I’m willing to give them the benefit of the doubt.

      No one who understands the basics of computer science uses the phrase “Turing problem”. There is the Turing Test, which is what they were talking about in this video and the Halting Problem which Turing developed the relevant proof for. The two are completely different and except that they both involve Turing and computers have nothing to do with each other.

      • In reply to #27 by Red Dog:

        No one who understands the basics of computer science uses the phrase “Turing problem”.

        OK, Turing Test then, was trying to be all philosophical :P

        Has anyone yet solved this problem though?

        I call it a problem because I work in a call centre and would like to see my job made redundant.

        • In reply to #28 by Peter Grant:

          In reply to #27 by Red Dog:

          No one who understands the basics of computer science uses the phrase “Turing problem”.

          OK, Turing Test then, was trying to be all philosophical :P

          Has anyone yet solved this problem though?

          It’s an important distinction. The Halting Problem is a theoretical finding in computer science about what you can and can’t prove about programs. So no one can “solve it” or rather Turing did solve it to the extent he developed the proof, until his work it was still an open question.

          The Turing Test is different. Note that in what follows I’m talking about the Turing Test as it’s typically wrongly defined, as Dennett did here, not the original test that Turing wrote but the popular conception that everyone uses: to make a computer program that can fool a human into thinking it is also human. That is as much an engineering question as a philosophical one and unlike the Halting Problem there isn’t a proof or even an unambiguous definition of what would constitute a solution. I’ve actually wondered myself why no ambitious AI researcher who likes headlines hasn’t tried to “solve it”. With current technology I think you could do really well. When the Watson team first started working I wondered why they didn’t just do a Turing Test rather than winning Jeopardy! although in hindsight I can see why they went the route they did. Even with the approach they took the reporting got dumbed down to “man vs. machine” and it would have been far worse if they had actually tried a real Turing Test.

          My guess is that is why no one has done it, the PR about trying to replace humans would be counter productive to the research goals and also there are better empirical tests and tasks that have more commercial applicability than a program designed to appear human would be. Doing well on Jeopardy! has enormous additional potential for general question answering and advice giving in all sorts of domains with complex unstructured information.

          • In reply to #29 by Red Dog:

            Note that in what follows I’m talking about the Turing Test as it’s typically wrongly defined, as Dennett did here, not the original test that Turing wrote but the popular conception that everyone uses: to make a computer program that can fool a human into thinking it is also human. That is as much an engineering question as a philosophical one and unlike the Halting Problem there isn’t a proof or even an unambiguous definition of what would constitute a solution. I’ve actually wondered myself why no ambitious AI researcher who likes headlines hasn’t tried to “solve it”. With current technology I think you could do really well.

            Well then please do, I’m getting sick of pretending to be human.

  12. A bit more on the Turing Test. One could claim that there already is good evidence that computers have passed it. One of the first rule based systems started out mostly as a goof by some AI researchers. They developed an AI system that was completely primitive by today’s standards and used simple pattern matching templates. It was a pretend psychotherapist. When it couldn’t parse the next statement in any meaningful way it just had a bunch of catch phrases it would use to hide the fact “tell me more”, “how do you feel about that”, etc. The thing that amazed the researchers was that college students were taking it seriously and were actually telling it their problems and considering the “advice” the program gave them.

    I always thought it said as much about psychotherapy as it did about AI. It’s also a good lesson on what it takes to make AI work in the real world. In my experience the sophisticated theorem provers often lose out to much less sophisticated approaches that just involve some clever programming.

    Sorry, I’m rambling, my point is that this doctor program was many decades ago and since with things like Siri there have been amazing advances and I think there are bots now that in limited domains pretty much do pass the test, that people interact with them as they would with a human. I know I’ve yelled at those damn voice recognition systems “I just want to talk to a F*ing person!!” more than once.

    • In reply to #31 by Red Dog:

      My main question is can they sell! I have recently moved from tech support in to the sales domain, mainly because it is more profitable, but I find it rather soul destroying. Even though I am rather good at it, I would be gratified if a computer could do it better.

  13. Antonio Damasio says that self aware consciousness can be defined as an exchange between the map of our own bodies, generated at the point where the spinal chord attaches to the brain (in the upper brain stem) and everything else is just additional complexity layered onto the base map. If this is true then consciousness can literally be defined as the focal point of physical experience. Plants, Jellyfish etc do not have this and so are not self aware, rats dogs and humans do and so they are. The reasoning is simple, conscious thought which is fluid and changeable needs a stabilizing base (our bodies, not our brains) in order to know what it is every morning when we wake up and go about the day.

Leave a Reply