From Schrödinger's Killer App, Chapter six.
To close this section, this chapter, and this book I want to talk about quantum biology and particularly the notion that the human brain already is a quantum computer. We must remember that Searle’s strong classical AI hypothesis is just that, a hypothesis. Hypotheses in science must be tested. I like to think it is true and I hope in 20 years or so, when our electronic computers have the same number of transistors and interconnects as our brain has neurons and synapses, we’ll find out. But if the classical computers continue to become more and more powerful without showing signs of self-awareness then perhaps other hypotheses should be considered. Quantum biology is a new area of research that postulates that what I will call the ‘weak’ quantum biology hypothesis. We have seen in this book that quantum physics, particularly quantum entanglement, offers an advantage over classical physics in computational power (on some problems) and in sensing and imaging (in some systems). Often, when nature offers a survival advantage that that biological life form will evolve to take advantage of it. Biological life forms take advantage of light sensors (eyes), sound sensors (ears), touch sensors (skin), and computational power (brains). The better you can see, hear, sense touch, or think the more likely you will be able to survive and pass on your genes for such things. The quantum biology hypothesis states that, since quantum mechanics offers advantages in sensing, imaging, and computing, biological systems should evolve quantum-based sub-systems to take advantage of those advantages. That is, according to the hypothesis, since biology takes advantage of any physiological edge offered to it, there should be biological life forms that already taking advantage of such uniquely quantum features as quantum unreality, uncertainty, and nonlocality.[i]
The US Defense Advanced Projects Agency even has a program in quantum biology. American physicist and DARPA program manager Matthew Goodman runs this program. When the program had its kickoff meeting in September of 2008, and somewhat to my puzzlement, Goodman invited me to attend. As I recall I talked to him on the phone and protested that I did not really know much about biology, quantum or otherwise, and that I was not working in the field. Did he really want me to attend and if so, pray tell, why? The answer was that I was to sit in front as skeptic-in-chief and use ‘the best bullshit detector in the business’ to advise him one what might be good avenues to pursue for research and what might be just a little nuts. I did not know much biology but I sure did know quantum mechanics. This was a role I aspired to. So I showed up in the DC area for the two-day workshop, sat in front next to Goodman, and proceeded to heckle all the speakers. Some of them became a bit irritated with me at first until Goodman and I explained my role and then they lightened up. In fact I think I was useful. Some of the talks seemed sound but some, to me, seemed to be nothing more than quantum numerology. After the conference I gave Goodman my advice on what seemed like good ideas to follow up on and what seemed just silly and every since I have kept one eye on the field of quantum biology. The two particular areas of current interest are in photosynthesis and bird migration. It appears in some photosynthetic bacteria that live in water deep enough to be dark the bacteria harvest photons with an efficiency that cannot, yet, be explained with classical theory alone. The photons arrive at an antenna-like ‘light harvesting’ structure in the photosynthetic bacteria and then with a very high probability, much higher than classical physics can explain, the photon energy is transported to a reaction chamber where it is converted into chemical energy to power the bacteria.[ii] My complaint about this claim, at least in 2008, was that the experiment that demonstrated the effect was carried out at the frigid temperature of liquid nitrogen, which is –196°C (–321°F). Those bacteria are not doing anything at such a temperature — they are frozen solid! The experiment was suggestive but certainly not conclusive. I recommended DARPA fund the experimenters to redo the experiment at room temperature. Biological organisms on Earth are not selecting for anything at –196°F; they are dead. The experiments have been done at room temperature and although the effect is not quite as startling but — much to my surprise — at least some great degree of quantum coherence, quantum unreality and cat-states, survives at room temperature.[iii] I’m puzzled by this result in that in the world of quantum technology often objects must be cooled to very cold temperatures for quantum coherence to survive. As things are heated up the thermally fluctuating environment should tend to destroy the coherence at room temperature. Perhaps nature has, over millions of years of evolution, found a way to protect quantum coherence in warm biological environments or mitigate the effects of the swirling thermal fluctuations in the hot soup of life. Why on Earth is DARPA interested in this stuff? Well if bacteria have found a way to make more efficient photon absorbers perhaps we can learn from them Rather than spend billions on making improved photon collectors for solar cells we just lift the technology out of the bacteria and place it solar panels on our roofs.
A second canonical example of what is suspected to be a quantum biological magnetic field sensor in the brains of migratory birds or more particularly their eyes. It has been known for thirty years that some birds use the very weak Earth’s magnetic field to navigate over trans-global distances. The problem is that all known mechanisms from classical mechanics and ordinary chemistry cannot explain the sensitivity of any biological magnetic field sensor that could do this. So after thirty years it is time to give the quantum biologists a chance. What is known is that the magnetic field sensor is activated when light hits the eye of the bird. The weak quantum biology hypothesis is that the evolutionary advantage to a migratory bird of having and Earth magnetic field sensor would be so strong that if any such mechanism ever arose by chance mutation the evolutionary amplification process of natural selection would size upon it and develop it into a quantum bio-technology that would benefit future generations of bird brains.
The model is that the photons striking the bird retina create a pair of spin-entangled electrons in a chemical reaction, and then those spins respond to the magnetic field with a signal-to-noise greater than say the spins of two uncorrelated electrons. One proposal is that when the pair is one of the four possible two-spin quantum states it produces a chemical that it does not when it is in one of the other three spin states. The strength and orientation of the field determines how many of the pairs are in the one versus three states, and so it is presumed, affects the rate at which the chemical is produced. Then somehow, it is not clear, perhaps the bird sees something it its eyes that corresponds to the magnetic field direction and strength and then uses this information to steer itself on its bi-annual migrations north or south. Again my concern is that quantum entanglement, and the required quantum coherence required to produce it, is very fragile and very quickly destroyed by the thermal fluctuations in the biological environment of a relatively hot living bird. But perhaps evolution has found a way to protect the entanglement that we have not. Evolution is a powerful thing. Or maybe the entanglement only needs to survive a few nanoseconds to do its job and produce the right ratio of chemicals that color the magnetic field across the bird’s field of view. Again, DARPA never met a magnetic field sensor it didn’t like. If we could reverse engineer this quantum bio-technology perhaps we could build room temperature super-sensitive magnetic field sensors that we could then integrate into a chip, the size of a grain of sand, and implant behind our ears to read our minds and allow our teenagers’ teenagers’ teenagers’ teenagers to carry out oblivious pseudo-telepathy with each other .
This leads me to what I will call the strong quantum AI biology hypothesis. This hypothesis has, which today has few followers, been most forcefully argued by British mathematical physicist Roger Penrose in his 1989 book The Emperor’s New Mind, and less forcefully argued by our old friend Henry Stapp in his 1993 tome, Mind, Matter, and Quantum Mechanics.[iv] The strong quantum AI biology hypothesis is in direct contradiction to Searle’s strong classical AI hypotheses, and it states that no appropriately programmed classical computer with the right inputs and outputs, no matter how powerful, will ever have a mind in exactly the same sense human beings have minds — that is human minds are fundamentally different than classical computers and that quantum mechanics is required to explain human consciousness. That is, the strong quantum AI biology hypothesis posits that the human mind is in fact already a quantum computer; that hundreds of thousands of years ago some quantum effect or effects arose by mutation in the mind of our ancestors and gave our brains a computational advantage over the rival progenitors running on meat processors only. As evolution is wont to do, this slight computational advantage was greatly amplified through the process of natural selection until it produced the end result; the human mind. The strong quantum AI biology hypothesis states that we have already met the sentient quantum computer and that he is us!
Stapp’s argument stems from his belief that that human mind routinely engages in ESP and that quantum entanglement is needed to explain ESP and so the human mind must be fundamentally quantum. As I have argued vociferously above, after 50 years of controlled experiments, there is absolutely no evidence for ESP and plenty of evidence against it. ESP does not exist and so there is no need to posit quantum mechanical processes in the brain to explain it. Stapp’s logic seems to me to be that he does not understand how quantum mechanics works, and he does not understand how ESP works, and so he argues that quantum mechanics is required to explain his how the mind engages in ESP. I dismiss this argument out of hand since ESP does not exist and so does not need explaining. Trickier is the argument of Penrose. Penrose simply rejects the strong classical AI hypothesis. That is Penrose declares, without any evidence to support his position, that no classical computer, no matter how powerful, can ever have a mind in the same way that a human has a mind. I have read his book and heard him talk on the subject and as far as I can tell his argument goes like this. Penrose does not understand how quantum mechanics works, and he does not understand how his brain works, and hypothesizes the quantum mechanics is needed to understand the working of the mind. I suspect Penrose just looks at his desktop PC and thinks that, There is no way that thing will ever be as smart as me!” To be fair there is no evidence for the strong classical AI hypothesis but we might want to rule it out first based on experiment before invoking the strong quantum AI biology hypothesis. Revulsion at the thought of your desktop PC someday having a mind equivalent to your own is not experimental evidence for rejecting the strong classical AI hypothesis. Those of us who watch the television series Star Trek not only find some appeal in machine minds but I daresay some of us even identify with such android humanoids such as Lt. Commander Data.
The problem that I have with Penrose’s strong quantum AI biology hypothesis is similar to that I had with the light harvesting bacteria and the magnetic bird sensor. Delicate features of quantum weirdness, unreality, uncertainty, and nonlocality, are easily destroyed by the thermal fluctuations of the environment, which are particularly severe in hot-blooded animals such as birds and humans. As American astrophysicist Carl Sagan was fond of saying, “Extraordinary claims require extraordinary evidence.” The claim that that some bacteria have evolutionary exploited weak quantum effects to make better photoreceptors or the claim that some birds have evolutionary exploited weak quantum entanglement to make better magnetic fields sensors are not extraordinary claims and so a few tight non-extraordinary experiments on the bacteria and the birds should be enough to prove this one way or another to my satisfaction. I will be a little surprised if the weak quantum biology hypothesis turns out to be true but when I am surprised I am happy. If these experiments pan out then it will be very interesting to learn what nature has done to protect these weak quantum effects from the thermal environment and indeed perhaps we can exploit what nature has done to make better photoreceptors and magnetic field sensors.
However Penrose’s strong quantum AI biology hypothesis is orders of magnitude more extraordinary and so the evidence to prove it needs to be orders of magnitude more extraordinary. It is a long way to go from nature having found a way to protect a few quantum states so that pigeons can migrate to nature has found a way to build a large-scale quantum computer in our noggins so that we can think. In The Emperor’s New Mind, Penrose offered no concrete model for just how quantum entanglement would lead to consciousness.[v] It was just a lot of wishful thinking and the reviling of the strong classical AI hypothesis. After falling into cahoots with the notorious anesthesiologist and hawker of quantum consciousness, Stuart Hameroff, Penrose published a 1994 book called Shadows of the Mind, where he proposed changing quantum theory to fit his hypothesis of quantum consciousness and further proposed that, without any evidence whatsoever, that there are ‘microtubules’ in the brain that somehow store and protect from thermal noise the fragile quantum entangled states purportedly needed to explain human consciousness.[vi] My bullshit detector simply pegged. Change the laws of quantum mechanics to fit your hypothesis? Postulate, with no evidence, microtubules in the brain to support your hypothesis? When you start changing the rules of the game to fit your pet hypothesis this then is the hallmark of pathological science.[vii] To summarize there is, in spite of a brief experimental search, no evidence of ‘microtubules’ in the brain that store quantum states, much less that the brain uses them as quantum processors to generate human consciousness.[viii] As far I can see there is also no reason at all to change the laws of quantum mechanics — the most successful theory of all time.
What is Penrose’s beef with the classical computer? From reading The Emperor’s New Mind it is difficult to tell as that book is all over the map. The reader is introduced a wildly disparate collection of topics such as Newtonian mechanics, quantum mechanics, cosmology, and quantum gravity before Penrose attacks the strong AI hypothesis in the last couple of chapters. No physicist in his or her right mind would think quantum gravity has anything to do human consciousness. Nor would any evolutionary biologist. Again it is one thing to posit that evolution has made better bacterium photoreceptor using bits quantum flotsam and jetsam and quite another thing to propose that evolution has harnessed the hypothesized quantum fluctuations of space and time in order to build a human mind. The human mind is about 10 centimeters across while quantum fluctuations in space are about 10–33 centimeters across. How on Earth would evolution in a series of gradual steps bridge those 32 orders of magnitude in distance to harness quantum gravity and why on Earth would it need too? Once you dig through mountains of chaff in his book you find one single kernel of barley. Writing in 1989, Penrose complains that all electronic computers of that age are classical universal computers in the Turing sense, which is equivalent to a classical Turing machine, and for that reason can never mimic the behavior of the human mind. He might have had a point in 1989 but since then all sorts of new classical computing paradigms have sprung up. Searle’s strong AI hypothesis states, “The appropriately programmed computer with the right inputs and outputs would thereby have a mind in exactly the same sense human beings have minds,” but never specifies that the computer must be algorithmically equivalent to a Turing machine. The claim that all electronic computers must be equivalent to ordinary Turing machines is Penrose’s own personal straw man, which he then merrily ignites with a blowtorch and then dances gleefully about the flames whilst lobbing Molotov cocktails in the pyre.
The reader may wonder why at this very late junction I have decided to hammer on Penrose and then Hameroff. Well aside from the point that the Rube Goldberg constructions, modified quantum theory and unseen microtubules, they require to provide a quantum basis for human consciousness are just silly, there is no need to invoke such a quantum basis, at least not yet. The idea that human consciousness is quantum based, Penrose’s strong quantum AI biology hypothesis, is an extraordinary claim. But in spite of all the smoke from his smoldering smudge pot of his burning straw man, I find very little in the way of flames. There is no evidence at all to support this claim much less the extraordinary evidence that Sagan would require. There is no reason yet at all to rule out Searle’s strong classical AI hypothesis. Penrose’s quixotic attack on the Turing machine model of computation is completely off base. The real surprise is that Turing’s simple model of computation has taken us as far as it has and not that it is the end of the story of classical computation. As I have related above, neuroscientists conjecture that human consciousness lies all in the synapses, the interconnects in the brain.
When Penrose penned his first book in 1989 on this topic, The Emperor’s New Mind, the science of artificial neural networks was mildly popular. Now it is wildly popular. Neural networks are models of very classical computing that are actually taken from models of the human brain. Lots of transistors, lots of interconnects, and lots of feedback loops. Some artificial neural networks are equivalent to Turing machines but others appear to be super-Turing, which is they have properties that transcend the simple computational model of computing Turing proposed eighty years ago. Penrose is attacking an eighty-year old model of computing. It would be a surprise if there had been no progress in classical computing since them. The field of super-Turing machines, while not without its own controversies, provides a framework where a neural network can carry out tasks not in the usual universal computing framework, the framework Penrose attacks. Particularly interesting is that super-Turing machines may have a reflexive or self-referential or highly recursive. That is super-Turing computers have a built in ability to think about themselves, a hallmark of human consciousness.[ix] Hofstadter in his 2007 book, I Am A Strange Loop, expounds on this idea an particularly makes the case that sufficiently complex but classical self-referential systems, such as possibly neural networks, will necessarily develop an illusion of self and therefore posses unique properties of a human mind. This is all not proved and worked out and itself constitutes and extraordinary claim, a sufficiently powerful artificial neural network will have a mind in the same way a human has a mind, but such a statement falls in the purview of the Searle’s strong AI hypothesis; Searle never claimed his powerful computer was a Turing machine, only Penrose claimed this. My point is that this neat set of ideas needs to be investigated and ruled in or ruled out before invoking quantum gravity or microtubules or whatnot to explain human consciousness. In 20 or 30 years we shall build (and perhaps eventually merge with) a powerful, self-referential, but still classical AI, a super-Turing AI, based on an artificial neural networks with a 100 billion transistors and a 100 trillion interconnects and we shall wait to see if it wakes up and talks to us and then passes the classical Turing test for consciousness. If it does not and then we continue onwards for 40 or 50 years with a trillion transistors and a quadrillion interconnects and still no sign of sentient life, well then we can start revisiting the strong classical AI hypothesis and perhaps reject it; but not now.
The reason I have spent so much time on Penrose’s proposal is that I think Searle’s strong classical AI hypothesis is right. Quantum mechanics need not be invoked to explain the human mind. I am, still in the end, proud to be a meat computer. But I also believe in my own strong quantum AI hypothesis, which I want to carefully peel away from Penrose’s strong quantum AI biology hypothesis. There will someday arise a quantum mind. The appropriately programmed and sufficiently powerful quantum computer, with the right inputs and outputs, have a mind in exactly the same sense human beings have minds, but it will have a mind that, unlike me, also thinks in Hilbert space and therefore super-exponentially transcends the human mind. If Penrose is right and my mind is a quantum computer, well then my mind is a particularly lousy quantum computer. I can immediately construct a question for the quantum Turing test that I myself cannot pass. Says the quantum mind to me, “Dowling! Can you factor this hundred-digit integer into its composite primes in under a second?” No, I confess, to it, I cannot. If Penrose is right and I am some sort of quantum computer then I am the crapola of all quantum computers. I make this point precisely so that when the true quantum technology based quantum mind comes online in a hundred years the acolytes of the Church of the Larger Penrose Space do not waive ancient tattered copies of his book about and declare victory — that Penrose was right all along and that consciousness does indeed require quantum theory. I would extol them to remember he only claimed that my meat computer, my mind, requires quantum theory and not that quantum mind that emerges from our quantum technologies in a hundred years. What will that quantum mind be? Well I can try to predict exponential growth but I dare not try to predict super-exponential growth. What will a self-replicating life form that thinks in Hilbert space be like? Well I don’t know but it will think in a fundamentally different way than I do. When it arises from our quantum technologies what we will it do to us, or what will we do to it?
But is your brain a quantum computer? Probably not.
[i] See, “Quantum Biology,” in Wikipedia (Wikimedia Foundation, 12 August 2012) <http://en.wikipedia.org/wiki/Quantum_biology>.
[ii] See, “Quantum Secrets of Photosynthesis Revealed,” by Lynn Yarris in Research News (Lawerence Berkely Lab, 12 April 2007) <http://www.lbl.gov/Science-Articles/Archive/PBD-quantum-secrets.html>.
[iii] See, “The Physics of Life: The Dawn of Quantum Biology,” by Phillip Ball in Nature, Volume 474 (2011) pages 272–274 <http://www.nature.com/news/2011/110615/full/474272a.html>.
[iv] See, The Emperor’s New Mind, by Roger Penrose (Oxford University Press, 1989) <http://www.worldcat.org/oclc/19724273>; and Mind, Matter, and Quantum Mechanics, by Henry P. Stapp (Springer, 1993) <http://www.worldcat.org/oclc/28147220>.
[v] See, “The Emperor’s New Mind,” in Wikipedia (Wikimedia Foundation, 13 August 2012) <http://en.wikipedia.org/wiki/The_Emperor%27s_New_Mind>.
[vi] See, “Is the Brain a Quantum Device?” by Victor Stenger in the Skeptical Inquirer, Volume 18.1 (2008) <http://www.csicop.org/sb/show/is_the_brain_a_quantum_device/>
[vii] See, “Pathological Science,” in Wikipedia (Wikimedia Foundation, 14 August 2012) < http://en.wikipedia.org/wiki/Pathological_science>.
[viii] See, “Shadows of the Mind,” in Wikipedia (Wikimedia Foundation, 13 August 2012) <http://en.wikipedia.org/wiki/Shadows_of_the_Mind:_A_Search_for_the_Missing_Science_of_Consciousness>.
[ix] See, “Hypercomputation,” in Wikipedia (Wikimedia Foundation, 13 August 2012) <http://en.wikipedia.org/wiki/Hypercomputation>.