catallaxy files

catallaxy in technical exile

Libertarians and the Singularity

with 70 comments

This is not really a proper review as I’m only up to Chapter 3 but I’m currently reading Ray Kurzweil’s The Singularity is Near and it’s probably the most exciting book I’ve read all year.

I’ve been waiting for a work like this, which provides an overview of the cluster of technological innovations that Transhumanists consider important, the science behind these innovations and a plausible assessment of their future pace of improvement.

Now if all there was to Kurzweil’s theory was invoking Moore’s Law and drawing logarithmic graphs, this would be a boring book. And admittedly Kurzweil is fond of drawing graphs. But obviously he does do more than that. He crunches the numbers, for example, to estimate the computational power that would be needed to simulate the workings of the human brain. One estimate is 100 trillion computations per second according to Hans Moravec who extrapolates from the computations per second needed for the neural image processing in the retina. Kurzweil argues plausibly based on how current implantable devices for controlling insulin levels work that the actual processing power would be less than the sum of that involved in simulating each neural cell since there would be no reason to. He concludes with a conservative estimate of 10 to the power of 16 cps. He then compares this estimate with the speed of progress achieved so far in computational power. Anyway the bottom line is he expects computers to pass the Turing test by the end of the 2020s.

Kurzweil also postulates that innovation occurs as an ongoing exponential sequence made up of a cascade of S curves (i.e. slow growth, rapid growth and plateau). Of course this is nothing new to people familiar with the economics of innovation. He even has a section on why growth deflation is a good thing (something that would warm the heart of our prolific commenter Bird). He’s clearly put a lot of thought and research into this book and I look forward to reading the extensive end notes which go for 103 pages and provide a wealth of references to chase up other research. As for the structure of the book, here are the chapters for the first half which would provide a good idea of the kinds of technologies that transhumanists consider important for achieving the singularity:

  • Chapter 1 – The six epochs
  • Chapter 2 – A theory of technological evolution
  • Chapter 3 – Achieving the computational capacity of the human brain
  • Chapter 4 – Achieving the software of human intelligence – how to reverse engineer the human brain
  • Chapter 5 – GNR (Genetics, Nanotechnology and Robotics)
  • The rest of the book is devoted to addressing critics.

    Singularity theory has two kinds of critics. The first kind of critic are those who doubt the plausibility of the kinds of innovations Kurzweil talks about. A good example is the philosopher John Searle who has proposed the famous Chinese room thought experiment to refute claims of strong AI (artificial intelligence). Count me in among those who believe that the mind is basically just what the brain does and if dualism is false (as it obviously should be seen to be to anyone who takes naturalism seriously) then there is no reason why consciousness cannot be simulated. Searle’s Chinese room thought experiment misses that consciousness doesn’t work on the principle of a little man sitting inside your head anyway.

    The second kind of critic is the one who thinks that even if all the things Kurzweil talks about is possible, it isn’t desirable. I notice for instance that in one of his chapters Kurzwel talks about the evil anti-human ideology called the precautionary principle. This second camp unites a lot of unlikely bedfellows – religious traditionalists who worry about us losing our ‘humanity’, secular neoconservatives like Leon Kass and lately some elements of the Left (actually an increasing number of elements of the Left). The latter development doesn’t surprise me. Insofar as a defining element of the Left is an interest in social justice (which I think it is) and essentially what the fetish with social justice amounts to is, as Hayek correctly pointed out, an atavistic desire for ‘society’ collectively to be so in control that it should even be concerned with freezing certain distributional patterns in time forever, then that is a clearly conservative instinct. This instinct has latterly been reinforced with their enthusiastic adoption of the ‘precautionary principle’, which as Don Arthur has pointed out is fundamentally incompatible with progress.

    For the record, I’m generally on the side of Kurzweil’s forces of optimism on both issues. I think the Singularity is desirable and the best way that libertarians can help hasten it is to ensure that governments get out of the way of locking us into or out of particular technologies. I think when it finally arrives, a lot of the debates we have today about things like the definition of the family and marriage, the pace of economic change and restructuring and outsourcing, ‘work-life balance’, what we can and can’t censor and, dare I say it, even global warming, may seem rather silly and irrelevant.

    Anyway, the point of this post was just to raise a number of topics for debate. I thought this book would be of interest as many libertarians hang around here and there seems to be a huge overlap between libertarianism, technological optimism and transhumanism. Any reason why this is so besides the obvious? It’s struck me that libertarians tend to be optimists generally and therefore people with an optimistic mindset tend to self-select into libertarianism and transhumanism.

    As a side note there doesn’t seem to be many Australians who write about transhumanism but two that I have heard of are Damien Broderick and Russell Blackford. Readers might also like to check out Blackford’s excellent essay on Genetics, ethics and the State which appeared in the usually vaguely fogeyish Quadrant.

    Advertisements

    Written by Admin

    November 26, 2006 at 2:21 pm

    Posted in Uncategorized

    70 Responses

    Subscribe to comments with RSS.

    1. “I think the Singularity is desirable and the best way that libertarians can help hasten it is to ensure that governments get out of the way of locking us into or out of particular technologies”

      But what is it?

      If its a single point in time where technological progress de-couples from capital accumulation and turnover…

      Well then I think its make-believe. But I’d want to know what I’m disagreeing with.

      Because what I’ve disagreed with there is like saying hot looking babes smiles will decouple from their faces.

      But probably the bulk of the rest of the things he’ll talk about I could likely agree with.

      GMB

      November 26, 2006 at 3:08 pm

    2. Jason Soon

      November 26, 2006 at 3:13 pm

    3. and this is from the blurb of his book:

      >>>>>
      in The Singularity Is Near, he examines the next step in this inexorable evolutionary process: the union of human and machine, in which the knowledge and skills embedded in our brains will be combined with the vastly greater capacity, speed, and knowledge-sharing ability of our own creations.

      That merging is the essence of the Singularity, an era in which our intelligence will become increasingly nonbiological and trillions of times more powerful than it is today—the dawning of a new civilization that will enable us to transcend our biological limitations and amplify our creativity. In this new world, there will be no clear distinction between human and machine, real reality and virtual reality. We will be able to assume different bodies and take on a range of personae at will. In practical terms, human aging and illness will be reversed; pollution will be stopped; world hunger and poverty will be solved. Nanotechnology will make it possible to create virtually any physical product using inexpensive information processes and will ultimately turn even death into a soluble problem.

      Jason Soon

      November 26, 2006 at 3:15 pm

    4. But there he’s talking about an ERA.

      Whereas a singularity is a moment-in-time.

      Its a threshold seperating things into the before and the after.

      Is he quits with the notion of this moment of explosive progress?

      Wonder if this singularity will burst forth in Darfur all of a sudden.

      See the rapidly evolving technology, to be sustainable, must sit at the top of the pyramid of more humdrum investment, maintenance, overtime, savings and general slogging away, trying to pay the bills and stuff.

      You don’t get the icing on the cake without the cake or the blade on the knife without the knife.

      He’s dead right about the growth deflation though.

      Because the worst waste of resources there is next to socialism outright, is when you have gotten this investment boom going, and then it morphs into a speculative boom.

      Its just sets everything back so many years or decades.

      GMB

      November 26, 2006 at 3:22 pm

    5. “In futures studies, a technological singularity (often the Singularity) is a predicted future EVENT believed to precede immense technological progress in an unprecedentedly brief time. ”

      Its just this millenarial EVENT busines that I call bullshit on.

      The rest of the ideas there are fine.

      But really its all to do with getting capitalism and capital accumulation up to a new level.

      GMB

      November 26, 2006 at 3:26 pm

    6. Birdy
      The one thing that leads me to think it is coming or at least some is that the cost of capital equipment continues to get cheaper.

      Think of the lead time it takes to get an orphan drug to market these days from the point of discovery. It takes about 10 years and 2 billion dolls.

      Seeing we have come close to mapping DNA, we could put all that into a computer and and then let digitization do all the work. Even clincal trials could be computerized. It would take a few hours to check out a drug…. even bring it to market.

      We could tailor make compounds to fit our DNA map.

      That’s an example of singularity.

      Finally we could arrive at them point where humanity becomes redundant in nice way. Machines could be better and live forever.

      I always thought of the way that humans would be replaced is through machines.

      The main focus is that we must do all we can to scupper the lefts and the other stasists attempt to stop it. If we stopped taxes on capital we could get there in 30 years. We may just make it to live another 20 years.

      JC.

      November 26, 2006 at 3:26 pm

    7. We I Say we could get there in 30 years. I mean we could start seeing the beginning of the beginning. I think it would be breathtaking to watch.

      ….major discovery in hours. GDP growth measured in weeks…… this could happen.

      JC.

      November 26, 2006 at 3:29 pm

    8. Jason:
      >Searle’s Chinese room thought experiment misses that consciousness doesn’t work on the principle of a little man sitting inside your head anyway.

      From another angle of attack, Frank Tipler has nicely demonstrated that Searle’s famous argument is as physically impossible as jumping to the moon.

      Daniel Barnes

      November 26, 2006 at 3:30 pm

    9. Yes, you make a great pooint about capitla accumulation etc.

      But singulariry doesn’t have to be a western attainment if we don’t want it. Already the europeans have given up the ghost by their desire to favour present day consumption over investment. Asians to their credit are not. So we could reach this point with Asian capital accummulation.

      The Chinese accumulate at a very high rate. if the commie regime got off their backs, they could be the ones getting there first.

      I think whoever gets there first wins.

      JC.

      November 26, 2006 at 3:34 pm

    10. The man who wrote the Chinese Room thought experiment is clearly not a linguist. I only speak European languages, and I still don’t see how this is possible.

      That said, I didn’t know when I was reading Iain Banks and watching Gattaca that I was learning about transhumanism. This post exposed me to something with which I was almost completely unfamiliar.

      It strikes me as plausible, I have to say.

      skepticlawyer

      November 26, 2006 at 3:38 pm

    11. Thanks to Jason’s link, I’ve just read a piece by Damien Broderick that I think is the most glorious bit of bad writing I’ve ever encountered. Go here.

      I hope his novels aren’t like that. Mind you, he’s an Australia Council grantee, which doesn’t bode well.

      skepticlawyer

      November 26, 2006 at 3:51 pm

    12. I think we have to go over again exactly what we are talking about here.

      Are we talking about ETERNAL YOUTH.

      I mean that would be good and is quite on the cards.

      But when you mention it people go all strange.

      We cannot bundle a whole bunch of ideas together, most of them worthy, and then call it the Singularity and then expect economics to stop right there.

      We want to be specific as to what it is we mean.

      Does it mean that each of us will be able to afford two or three robot servants, smarter then us, to do our bidding?

      We want to look at some of these predictions seperately.

      GMB

      November 26, 2006 at 4:08 pm

    13. I think it means that a large amount of captial input and human knowledge will eventually make us eternal even if we/re living in a cyerspace.

      But I agree economics cannot be left out.

      JC.

      November 26, 2006 at 4:16 pm

    14. it’s a whole cluster of innovations coming on at the same time but I think the one that ties them all together is AI. Reality doesn’t have to be simple and cut into neat portions.

      Robot servants are passe, Graeme. You’re thinking small. They’ve already created female androids in Japan
      http://news.bbc.co.uk/2/hi/science/nature/4714135.stm

      If say we combine AI which passes the Turing test which Kurzweil think will come at the end of the 2020s, and we dramatically boost computing power through a breakthrough in quantum computing and we throw nano technology into the mix (computers self-assembled from nano-dust?), the possibilities are endless. we could have super-small super-computers with the power to simulate human consciousness. we could have different versions of our consciousness floating around, for one thing. We could have intelligent nano-bots working at a sub molecular level which can assemble whatever we want in the blink of an eye.

      Jason Soon

      November 26, 2006 at 5:10 pm

    15. I have one of Broderick’s books somewhere called The Spike which is non fiction and is an early Singularity book.

      I used to reasonably active on the Extropians list and he was on it, one of the few Australians. Also most extropians are libertarian though there is a left wing version and Damien is sorta left of centre so he stood out too. But the more such people the better.

      But the problem with the left is they’re always too concerned about ‘unequal access’ and crap like that so they would rather stop these things from happening or bring them under government control than allow it unleashed even if the end result is nothing happens. And if they’re not concerned about unequal access there’s the precautionary principle i.e. nature will be destroyed if we blend the peanut gene with wheat or something like that. Thus, more power to people like Damien.

      Jason Soon

      November 26, 2006 at 7:19 pm

    16. Huh! I notice that Damien mentions this on his website too on that link you pointed out, SL:
      >>>
      I met Barbara when we were both members of the global on-line community of technological optimists and intense individualists associated with transhumanism, an emerging philosophy with some overlap with programming, cutting edge science, and sf. There were more die-hard fans of Ayn Rand there than I felt comfortable with—being an old sixties communitarian-anarchist, to give something as squishy as my social philosophy a label. Most were Americans, and most were devoted to their nation’s right to bear as many private and lethal arms as possible, something that seems very alien and disturbing to an Australian. None of this turned me away, because what I shared with them was more urgent, significant and often hilarious than what sets us apart. But I feel more at home with certain extropians and transhumanists than with others.

      Jason Soon

      November 26, 2006 at 7:28 pm

    17. All of which he could have said in three (comparatively short) sentences.

      Note the presuming to speak for all Australians, too.

      skepticlawyer

      November 26, 2006 at 7:40 pm

    18. “I think it means that a large amount of captial input and human knowledge will eventually.”

      Eventually?

      You see this is the problem I have with it. Why not talk about these clusters of innovation that might make life better and longer?

      And talk about the transhumanism and the defeat of ageing?

      Bringing in the S-word just confuses matters.

      Capital goods not only take a lot of effort to develop. They also take a lot to maintain.

      The high-tech stuff only becomes cheap if its mass-marketable.

      And the cheap high-tech goods produced sit on this massive pyramid of capital and interlinked organisations all of which have to work profitably to be maintained and enhanced.

      These things seem to get left out of all these projections.

      What if I was to say we ought to be aiming at 20% plus growth in GDR?

      People would scoff at me but isn’t such a more practical and direct way of looking at things then in terms of this Singularity business?

      GMB

      November 26, 2006 at 9:33 pm

    19. “I think when it finally arrives, a lot of the debates we have today … may seem rather silly and irrelevant.”

      Yes if we achieve the end of civilisation, they will be irrelevant.

      I have no clue what this singularity theory is, but if singularity is understood in mathematical sense, then it is nothing but a global catastrophe.

      Sorry for the ignorant comment. It is 2 am here.

      Boris

      November 27, 2006 at 4:52 am

    20. I reviewed this a while back

      John Quiggin

      November 27, 2006 at 7:49 am

    21. Nice review, JQ – thanks for stopping by. Liked the bit about the slow boat to Australia in order for you to get your review copy 😉

      Certainly some technological changes are very rapid, and amazing – anyone over 25 who’s had a young relative ask them if they were allowed to take their mobile phone to school can testify to that – but like all progress, I think it’s likely to be uneven.

      Some parts of the economy will benefit, some won’t, and some will fall in the middle.

      skepticlawyer

      November 27, 2006 at 8:29 am

    22. I should read the book before I criticise, but I’ll have to say I’m skeptical. Its all very well to extrapolate say computing power, but even if we believe that we can keep following Moore’s law will the impact of that additional computing power have diminishing returns past a point?

      I question how much of the internet wizzbangery of the last few years is a result of increased processor power rather than people just realising new and effectively using the power we have, and advantages the increasing connection between them brings. ie. if proccessor speed increases stalled now would we notice any difference to the new net stuff coming out over the next few years.

      As for the point about AI why I do believe that intelligence is ultimately acheivable by computers, I also question whether increasing power rather say structure and organization is the key point.

      So that said I’m very upbeat about technological progress and what it will bring. However, I’m skeptical that it will occur in some sort of singularity, rather than a cointinuing series of breakthroughs, and particularly skeptical that we can place a meaningful date on such an event.

      Steve Edney

      November 27, 2006 at 9:23 am

    23. SE:
      >I should read the book before I criticise, but I’ll have to say I’m skeptical. Its all very well to extrapolate say computing power, but even if we believe that we can keep following Moore’s law will the impact of that additional computing power have diminishing returns past a point?

      I look forward to reading it too. But I agree with Steve’s tempered skepticism – isn’t AI always 20 years away? Plus isn’t the big hangup in AI ‘common sense’, which isn’t solved by exploding deductive power? Perhaps the book addresses these issues. I tend to agree with Popper, in that the problem of artificial intelligence will be solved after we’ve found a way of creating artificial life.

      Daniel Barnes

      November 27, 2006 at 9:59 am

    24. “The average American spends more time in the car, just to cover the basic tasks of shopping and getting to work, than was needed a generation ago, and in many cases, travels more slowly.
      The advocates of the Singularity tend either to ignore these facts or to brush them aside.” (JQ)

      Notice that this fact cannot be divorced from road socialism.

      If we look at this as an acceleration in the potential of our economy and the ability to extend and enhance our lives… If we do this and put the S-word away.

      Then we get a prescription for all the things we need to get done to take advantage of such an acceleration.

      GMB

      November 28, 2006 at 1:07 am

    25. “The average American spends more time in the car, just to cover the basic tasks of shopping and getting to work, than was needed a generation ago, and in many cases, travels more slowly.”

      I am not sure a decade ago, but everyone spends more time in the car than 100 years ago. But what does this mean?

      Boris

      November 28, 2006 at 2:25 am

    26. Im not sure of the generation ago thing also.

      But I suspect its about congestion due to no good pricing mechanism for roads, particularly at peak hour.

      Its easier to see it when it comes down to individual cities.

      Like I heard the horse-drawn vehicles moved more quickly, a century ago, then cars in Manhattan do now (or did recently).

      If true this is appalling inefficiency.

      GMB

      November 28, 2006 at 5:27 am

    27. Anyway the bottom line is he expects computers to pass the Turing test by the end of the 2020s

      Roger Penrose in his book “The Emperors New Mind” made the argument that a turing machine can never do some things that humans clearly can do. Eg A turing machine can not do predicate calculus but humans can. He stated (or showed) that this can be proven mathematically. I remember being pretty convinced by the book, although it was more than ten years ago that I read it.

      terjepetersen

      December 4, 2006 at 1:39 pm

    28. Like I heard the horse-drawn vehicles moved more quickly, a century ago, then cars in Manhattan do now (or did recently).

      If true this is appalling inefficiency.

      What about operating cost and the number of people moved per hour for a given area of road. I doubt that on these measures horses would be a more efficient option.

      terjepetersen

      December 4, 2006 at 1:45 pm

    29. Terje
      Artificial minds need not be based on exactly the same principles that current computers are based on. They are already developing artificial neural networks.
      I would be very sceptical of his claim. I don’t see how you could prove such a claim. I am guessing this is based on the claim that computers can’t understand Godel’s Theorem. How come we can? What’s so special about us that can’t be recreated if we are still made of elements from the natural world (as any artificial intelligence would be).,

      Jason Soon

      December 4, 2006 at 1:45 pm

    30. Artificial minds need not be based on exactly the same principles that current computers are based on.

      Correct. However turing machines are by definition based on the same principles as current computers.

      terjepetersen

      December 4, 2006 at 1:46 pm

    31. How come we can? What’s so special about us that can’t be recreated if we are still made of elements from the natural world (as any artificial intelligence would be).,

      This was the mystery that the book focused on. His theory was that there is quantum computing or some such thing happening in our brains.

      terjepetersen

      December 4, 2006 at 1:47 pm

    32. If you believe the world is a type of complex newtonian billiard game where all the bits bounce around deterministically then you are left without any scope for free will.

      terjepetersen

      December 4, 2006 at 1:49 pm

    33. Terje
      Everything is made up of quantum particles, including inanimate matter! He may be an accomplished physicist but I don’t think he’s on the right track here.

      I don’t believe in free will. It is like believing in God. I have covered this in other posts.

      Jason Soon

      December 4, 2006 at 1:53 pm

    34. Terje
      Passing the Turing test is just a technical phrase for ‘passing for human’. It doesn’t mean artificial minds will need to be based on the same principles as computers today.

      Jason Soon

      December 4, 2006 at 1:54 pm

    35. I don’t believe in free will.

      Your just saying that. 🙂

      Everything is made up of quantum particles, including inanimate matter!

      The point is that turing machines are like complex deterministic newtonian billiard games where as life and nature in general isn’t. The fact that inanimate matter is also made of quantum particles is beside the point.

      Determinism requires causation. And Einstein showed that causation and in particular event sequence is relative to the observer.

      With both general relativity and quantum physics giving special status to the subjective nature of the observer I think that an object reality is more suspect than free will. Although I doubt we will know for sure until the two theories are more fully reconsilled.

      terjepetersen

      December 4, 2006 at 2:01 pm

    36. That is not a very good interpretation of Einstein, Terje. And you’ve just swallowed a very New Age interpretation of quantum physics.

      Jason Soon

      December 4, 2006 at 2:03 pm

    37. Passing the Turing test is just a technical phrase for ‘passing for human’.

      I agree. However you can make a turing machine a gazillion times faster and give it a gazillion times more memory and it still remains a turing machine. If a turing machine can’t pass the turing test by mathematical definition then building a bigger turing machine is not the answer.

      terjepetersen

      December 4, 2006 at 2:04 pm

    38. If true then the following is pointless:-

      He crunches the numbers, for example, to estimate the computational power that would be needed to simulate the workings of the human brain.

      terjepetersen

      December 4, 2006 at 2:05 pm

    39. And you’ve just swallowed a very New Age interpretation of quantum physics.

      You would have to define “new age” before I could offer any meaningful reply. I’m not talking age of aquarius or any such thing.

      terjepetersen

      December 4, 2006 at 2:09 pm

    40. Roger Penrose may be wrong however on the face of it he does not appear to be a lightweight slacker:-

      http://en.wikipedia.org/wiki/Roger_Penrose

      terjepetersen

      December 4, 2006 at 2:13 pm

    41. Terje
      FFS
      1) I didn’t concede the point – I merely said they didn’t need to be based on the same design
      2) Doing things thru neural networks i.e. a differnet design (which is what our meat based computers in our heads are based on) doesn’t mean computational power can’t be measured.

      Obviously he had to calculate the computational power of our brain first to do the estimates.

      You are going half arsed spouting New Age babble

      Jason Soon

      December 4, 2006 at 2:14 pm

    42. Terje
      If you can provide the ‘proof’ in simple language which shows you understand it then do so. Otherwise you’re just going on authority. All those people working on AI obviously don’t agree that his ‘proof’ works.

      Jason Soon

      December 4, 2006 at 2:15 pm

    43. Let me put this again.
      Computational power can be measured regardless of the design of the machine whether it is based on parallel processing or otherwise.

      Jason Soon

      December 4, 2006 at 2:17 pm

    44. “turing machines are like complex deterministic newtonian billiard games where as life and nature in general isn’t”

      what exactly does this mean Terje? You are confusing ‘deterministic’ with ‘determined’.

      Financial markets and cloud formations are not deterministic. That doesn’t mean they are ‘undetermined’.

      Jason Soon

      December 4, 2006 at 2:25 pm

    45. Doing things thru neural networks i.e. a differnet design (which is what our meat based computers in our heads are based on) doesn’t mean computational power can’t be measured.

      I agree. However software based neural networks that run on turing machines are not going to get around the limitations of turing machines. And neither is a neural network that is functionally equivalent to one that can be simulated with 100% accuracy using software that runs on a turing machine. However there is scope for a physical design that overcomes any turing machine limitations, the trick is to discover what that design is.

      If you can provide the ‘proof’ in simple language which shows you understand it then do so. Otherwise you’re just going on authority.

      I can’t currently do it. No argument. I don’t profess to being an expert. However when I did take the time to look at this issue in the early 1990s I reached the conclusion that I now hold. I have since kicked away the logic ladder that got me here. So I am quite open to the possiblity that my earlier thinking processes were flawed and if I was going to go back and rethink the issue I might reach new conclusions. However I currently doubt that. I’m really just kicking the tires to see if there is anything knew on offer.

      All those people working on AI obviously don’t agree that his ‘proof’ works.

      Now it seems that you are making a counter argument that is based on authority (ie the operating assumptions of the majority).

      You are going half arsed spouting New Age babble

      You are going half arsed getting all in a spin because I don’t sing hallelujah and agree with you.

      You haven’t proved that fast turing machines can do what humans do. And other than saying “neural networks” you haven’t proven that AI specialists have any real clue about a design that can.

      terjepetersen

      December 4, 2006 at 2:29 pm

    46. terje
      see my comment 44
      your whole argument seems to be based on invoking some strawman idea of ‘Newtonian deterministic billiard balls’ that has been disposed of long ago. There is nothing magical about a system not being deterministic. That doesn’t make it alive or sentient.

      Jason Soon

      December 4, 2006 at 2:33 pm

    47. “turing machines are like complex deterministic newtonian billiard games where as life and nature in general isn’t”

      what exactly does this mean Terje? You are confusing ‘deterministic’ with ‘determined’.

      No I don’t think so. The halting problem shows that many turing machine outcomes are not determined until they happen. Deterministic as in the next state of the system is dependent entirely on the current state and the current inputs (and is repeatable if you can get the system to re-enter the same initial state and then apply the same inputs again).

      Quantum experiments such as the double slit experiment are not deterministic. They are probabilistic.

      terjepetersen

      December 4, 2006 at 2:38 pm

    48. There are many number sequence generating algorithms that are derterministic but not determined. In other words there is no short cut to finding the seqeuence of outputs other than by undertaking each and every step of the algorithm.

      I believe that knowing the N’th digit of pi is such a problem.

      terjepetersen

      December 4, 2006 at 2:42 pm

    49. Terje
      Probabilistic at the subatomic level doesn’t mean probabilistic at the macro level. If you’re trying to conjure free will out of this it ain’t gonna work.

      Jason Soon

      December 4, 2006 at 2:45 pm

    50. Probabilistic at the subatomic level doesn’t mean probabilistic at the macro level.

      Actually it does. We can arrange the experiment such that the end point the electron (after it passes through the double slit) will trigger some macro event (ie turning on the red light or else turning on the green light). The non deterministic nature of the double slit experiment is not due to randomness or inaccuracy in the setup. It is a real physical phenomena in which the next state of the universe is independent of the current state of the universe.

      terjepetersen

      December 4, 2006 at 2:51 pm

    51. next state of the universe is independent of the current state of the universe

      Independent is the wrong word. Not deterministic is the correct word but I was trying to avoid a circular definition. Essentially the experiment does not yield the same result each time but rather a probabalistic distribution of results.

      terjepetersen

      December 4, 2006 at 2:52 pm

    52. By the way I stopped believing in free in my late teens and only returned to the idea after I studied quantum physics.

      terjepetersen

      December 4, 2006 at 2:54 pm

    53. “free” should be “free will”

      terjepetersen

      December 4, 2006 at 2:55 pm

    54. And besides who said that “free will” was a macro scale process anyway. It may well be a micro scale process. Quantum computers certainly depend on micro processes to do what they do.

      terjepetersen

      December 4, 2006 at 2:57 pm

    55. I’m familiar with Schrodinger’s Cat Terje. We *can* arrange such a thing but so what?

      Going back to my earlier point, quantum processes are at work at a sub atomic level in a rock as well.

      Jason Soon

      December 4, 2006 at 3:02 pm

    56. Going back to my earlier point, quantum processes are at work at a sub atomic level in a rock as well.

      And in stars and plastic frizbies also. What is the point?

      terjepetersen

      December 4, 2006 at 3:04 pm

    57. Terje
      are you now conceding that quantum computers can solve the AI problem? Quantum computers are discussed by Kurzweil and are one of the fronts that AI researchers are working on. The fact that they are quantum computers doesn’t make their processing power irrelevant.

      So where does this leave Penrose’s proof?

      Jason Soon

      December 4, 2006 at 3:05 pm

    58. I was under the impression that Penrose’s view of the links between quantum physics and the brain were pretty much only believed by himself. Hawkins has made some pretty strong theoretical criticisms of his theories.

      Ken Miles

      December 4, 2006 at 3:06 pm

    59. Terje,you wrote earlier
      “His theory was that there is quantum computing or some such thing happening in our brains”

      My point is that there are quantum processes going on inside rocks and frisbees too. Calling it ‘quantum computing’ when you are talking about the meat machine inside our heads doesn’t distinguish the two.

      Jason Soon

      December 4, 2006 at 3:07 pm

    60. While I consider myself to be a technology optimist, I’m very very very sceptical about the singularity (that’s the polite version). I’m not impressed with the arguments put forward by transhumanists in popular culture (and Damien Broderick has the honour of writing the worst book which I have ever read – The Last Mortal Generation).

      When I compare claims about nanotechnology with the actual research (which is generally rebadged chemistry), there is a massive gap between the claims and hypothetical ideas in the scientific literature.

      Ken Miles

      December 4, 2006 at 3:17 pm

    61. Jason,

      Penrose was asserting that our brains are quantum computers (asserting not proving). He was asserting that it will takes a change of design to computer technology (ie machine hardware) to achieve intelligence in a machine. He was stating (ie offering a proof) that speed and memory alone would not turn a turing machine into an intelligent entity with human like skills.

      Quantum computers are not turing machines. They are fundamentally different in design. So yes I will happily concede that the door is open for quantum computers to achieve true intelligence in a machine (and I never meant to imply otherwise).

      However a shift from turing machines to quantum computers is not a logical outflow of moores law. Moores law just describes the progress of technology in the realm of turning machines. It is about faster and bigger turing machines, not a new design of machine. Moores law says nothing about breakthroughs that may or may not happen as a result of a fundamental reinvention of computing.

      In essence Penrose is saying that solving an old “class of problem” faster will not lead to intelligence. He is saying that intelligence requires the ability to solve a new class of problems that were unsolvable using the old turing machine design.

      Regards,
      Terje.

      P.S. The PC on your desk, the super computers that weather forcasters use etc are all turing machines.

      terjepetersen

      December 4, 2006 at 3:19 pm

    62. Jason, I think (and it has been a very long time since I last read up on the subject) that Penrose hypothesises that in order to examine how the brain works, you need to add some quantum effects. He suggests that these take place in tube like proteins in the brain. Stephen Hawking points out (in the same book – The Large, the Small and the Human Mind) that decoherence will overwhelm the quantum effects long before they could play a significant role in brain activity.

      I don’t know too much about this, but I passed it by a friend who was studing for his PhD in quantum physics and he agreed with Hawking.

      Ken Miles

      December 4, 2006 at 3:25 pm

    63. Thanks Ken. that was my rough understanding and that was why I was saying to Terje that I didn’t think quantum effects were relevant above the subatomic level. Otherwise we could be walking thru walls.

      Jason Soon

      December 4, 2006 at 3:27 pm

    64. Terje
      If we can get c8to to join this thread you will have a lot of fun. Not only is he a quantum skeptic, he’s a relativity skeptic.

      Jason Soon

      December 4, 2006 at 3:29 pm

    65. My point is that there are quantum processes going on inside rocks and frisbees too. Calling it ‘quantum computing’ when you are talking about the meat machine inside our heads doesn’t distinguish the two.

      The distinction is between “quantum computing” and “quantum processes”. There are quantum processses happening inside my PC (ie electrons crossing PN junctions on transisters that are etched on an IC) however that does not make my PC a “quantum computer”. Perhaps I was taking the term “quantum computer” for granted.

      Do we need to take a step back and discuss the distinctions that define a quantum computer?

      terjepetersen

      December 4, 2006 at 3:30 pm

    66. Otherwise we could be walking thru walls.

      You mean you haven’t mastered that yet? Obviously your just a beginner. 🙂

      terjepetersen

      December 4, 2006 at 3:31 pm

    67. In order to walk through walls you need to excite the brain atoms. This is best done with high doses of alcohol. Perhaps I can demonstrate it to you some time soon. You do need to be in a receptive state of mind however.

      terjepetersen

      December 4, 2006 at 3:34 pm

    68. yes I do know how quantum computers work, I just thought you and Penrose were begging the question. As you put it, he has only asserted that our brains work like quantum computers. And to be honest I’m not that convinced that quantum computers are necessarily qualitatively different.

      I may do a follow up post on all this stuff when I have time.

      Jason Soon

      December 4, 2006 at 3:40 pm

    69. And to be honest I’m not that convinced that quantum computers are necessarily qualitatively different.

      I suspect that Penrose has a pretty good grasp on how they work although I’m certainly a bit out of the loop these days.

      I agree that the current breed of quantum computers may not turn out to be qualitatively different to turing machines.

      The broader issue of whether computational systems (ie human brains, turing machines, quantum computers) are equivalent is an open question:-

      http://en.wikipedia.org/wiki/Church-Turing_thesis

      terjepetersen

      December 4, 2006 at 4:44 pm

    70. That last comment by me makes no sence. I think I’ll sleep on this for a while before I try again.

      terjepetersen

      December 4, 2006 at 5:03 pm


    Leave a Reply

    Fill in your details below or click an icon to log in:

    WordPress.com Logo

    You are commenting using your WordPress.com account. Log Out / Change )

    Twitter picture

    You are commenting using your Twitter account. Log Out / Change )

    Facebook photo

    You are commenting using your Facebook account. Log Out / Change )

    Google+ photo

    You are commenting using your Google+ account. Log Out / Change )

    Connecting to %s

    %d bloggers like this: