(For instance, a baseball in some sense needs to be modeled as a quantum system -- in the sense that the way its molecules hold together can be described only using quantum not classical physics; but classical physics can be used to explain the normally relevant aspects of its macroscopic behavior. So at the levels directly relevant for analysis of a baseball game, a baseball is best modeled as a classical system. OTOH, at the levels directly relevant for analysis of the electromagnetic behavior of a Superconducting Quantum Interference Device (SQUID) -- a small but macroscopic device, used in magnetoencephalography machines, and demonstrating macroscopic quantum coherence in its magnetic field -- the SQUID is best modeled as a quantum system. Classical physics models just won't explain why the SQUID, a device you can hold and pinch between your fingers (though it only works when supercooled, which would freeze your fingers!), makes MEG machines work.)
Current brain theory indicates that for understanding its role in giving rise to the mind, the brain is most effectively modeled as a classical system (i.e. the brain is more like a baseball than a SQUID) ... but of course current brain theory could be incomplete.
(Even if the brain is a macroscopic quantum system, this of course doesn't prove that quantum dynamics are necessary for intelligence or consciousness or anything like that. Those are bigger and deeper questions, and I've argued in the past that sufficiently complex "classical" systems might need to be treated using quantum logic ... but this gets into a lot of deep issues that I don't want to digress onto here.)
Stuart Hameroff is one of the more vocal proponents of the "quantum brain" idea, and he has a new paper reporting a new theory in this direction, arguing that dendro-dendritic synapses are mediated via macroscopic quantum dynamics, thus posing a quantum neural net that operates in complex coordination with the classical neural net formed by axonal-dendritic synapses.
I don't have a strong opinion on that particular theory of Hameroff's. I look forward to discussing it with him at the Toward a Science of Consciousness conference in Hong Kong next month.
But I was struck by one of the references at the end of his paper, a Nature paper entitled
Evidence for wavelike energy transfer through quantum coherence in photosynthetic systemsThis is a 2007 paper that I had not noticed before, and it's interesting because it gives solid evidence of macroscopic quantum coherence in a biological process.
To quote part of the abstract:
Here we extend previous two-dimensional electronic spectroscopy investigations of the FMO bacteriochlorophyll complex, and obtain direct evidence for remarkably long-lived electronic quantum coherence playing an important part in energy transfer processes within this system. The quantum coherence manifests itself in characteristic, directly observable quantum beating signals among the excitons within the Chlorobium tepidum FMO complex at 77 K. This wavelike characteristic of the energy transfer within the photosynthetic complex can explain its extreme efficiency, in that it allows the complexes to sample vast areas of phase space to find the most efficient path.
In the comments to an earlier edit of this blog post, someone pointed out this more recent paper
Quantum Zeno Effect Underpinning the Radical-Ion-Pair Mechanism of Avian Magnetoreception
whose abstract says
The intricate biochemical processes underlying avian magnetoreception, the sensory ability of migratory birds to navigate using earths magnetic field, have been narrowed down to spin-dependent recombination of radical-ion pairs to be found in avian species retinal proteins. The avian magnetic field detection is governed by the interplay between magnetic interactions of the radicals unpaired electrons and the radicals recombination dynamics. Critical to this mechanism is the long lifetime of the radical-pair spin coherence, so that the weak geomagnetic field will have a chance to signal its presence. It is here shown that a fundamental quantum phenomenon, the quantum Zeno effect, is at the basis of the radical-ion-pair magnetoreception mechanism. The quantum Zeno effect naturally leads to long spin coherence lifetimes, without any constraints on the systems physical parameters, ensuring the robustness of this sensory mechanism. Basic experimental observations regarding avian magnetic sensitivity are seamlessly derived. These include the magnetic sensitivity functional window and the heading error of oriented bird ensembles, which so far evaded theoretical justification. The findings presented here could be highly relevant to similar mechanisms at work in photosynthetic reactions. They also trigger fundamental questions about the evolutionary mechanisms that enabled avian species to make optimal use of quantum measurement laws.
This of course is even more intriguing than the green sulphur bacteria stuff, because it has to do with perception in an intelligent macroscopic animal.
Hameroff's point in citing the paper on green sulphur bacteria (and it's a good one) seems to be: if long-lived quantum coherence can play an important role in photosynthesis, couldn't it also play a role in the brain somehow ... e.g. maybe via dendro-dendritic synaptic gap junctions?
The extrapolation from these other results to neuroscience is speculative, sure.... But this kind of result does make the possibility of quantum coherence impacting human cognition seem a bit less fanciful.
After all, I often recall that in the late 90's all the neuroscientists I talked to told me there was no neurogenesis nor synaptogenesis in adult mammals. Oops. Now they've got new data and changed their mind. My point isn't that quantum coherence is related to neuro or synapto genesis (though, who knows...), but rather that neuroscientists -- simultaneously with displaying the usual humility of biologists regarding the complexity of the systems they're studying -- have a long-standing habit of assuming the concept-set underlying their current understanding is much more adequate than it really is.
Our ignorance of the brain is why my own AI work is not based on trying to closely model the brain. Of course, it's possible that intelligence is fundamentally based on some freaky neuroquantum phenomenon, so that all digital-computer AI work is doomed by some intrinsic limitations ... but I doubt it. My own guess is that, even if the brain does involve macroscopic quantum coherence in some interesting sense, one can still make transhumanly intelligent systems using digital computers. And of course, if this doesn't work -- or if these transhumanly intelligent systems turn out to lack some crucial aspect of self-awareness as the quantum-consciousness advocates argue -- then we can always add some funky quantum computing chips into our AGI server farm!
The "quantum module" is something I've considered in case we find that free will and self-awareness hinge on some fuzzy effect. Ultimately it may prove implementable in software
I'm really tired of the "mind is mysterious, quantum mechanics is mysterious => quantum mind" arguement.
Or, more specifically, "I feel like I have free will, quantum mechanics is non-deterministic => quantum mind".
Quantum mechanics isn't magic-- it's linear algebra. And as for the determinism issue, there's a difference between randomness and free will. Plus, there's a growing number of physicists who subscribe to the many-worlds interpretation of QM, which is purely deterministic.
Assuming for the moment that (a) quantum effects in neural functioning are necessary for human-level intelligence including AI, and/or (b) quantum effects are necessary for something like self-awareness, then a "quantum module" may be required for AGI and for mind-uploading (brain emulation). Could quantum indeterminacy be implemented in software alone (i.e. without quantum computing hardware)?
Michael ... Here's what's interesting...
Scientific data consists of finite sets of finite-precision observations. I.e., the total set of all scientific data can be packed into some set of bits in some digital computer program.
It follows that one could create a simulation world so that, to the residents of that simulation world, it appeared *based on all their scientific data* that they lived in a universe governed by quantum theory.
Yet, we would know that they "really" lived in a digital computer program.
The message: insofar as data-driven science goes, we can never refute the hypothesis that minds can be created inside some digital computer program.
But of course, it's possible that science is a fundamentally limited way to view the universe (due to its basis in finite sets of finite-precision observations), and that minds are not implementable using digital computers ... or quantum computers...
Whether it's a FEASIBLE approach to create minds in digital computers (versus quantum computers) is a whole other question, of course...
I happen to think it IS going to be feasible.... But I'm open to the possibility that it isn't.
If (though I doubt this is true) it turns out for some reason that mind can't be created in digital computers, then at very least AGI projects like mine will lead to amazing nonsentient technology to aid scientific discovery, create beneficial technologies, interface with human brains, and lead us toward whatever future tech WILL succeed in creating artificial minds...
"...the total set of all scientific data can be packed into some set of bits in some digital computer program."
Nice Ben. At the end of the day everything has to be computable.
Penrose (Hameroff?) argues failure of computable mathematical proof implies failure of computability. QM is a convenient mystery to take up the slack. But who canonized mathematical proof? In fact it's more reasonable to forget mathematical proof, and keep computability.
On the other hand as you say 'sufficiently complex "classical" systems might need to be treated using quantum logic'.
That's OK. Actually I think that abandoning the need for mathematical proof and keeping computability implies this.
Rob: I don't think I agree that "At the end of the day, everything has to be computable"
What I would say is rather: "At the end of the day, from the perspective of science or language, everything MIGHT AS WELL be computable."
Logically: This could mean that everthing is computable ... or it could indicate a fundamental limitation on the part of science and language relative to the universe...
"Logically: This could mean that everthing is computable ... or it could indicate a fundamental limitation on the part of science and language relative to the universe..."
I guess there is always that choice!
At the risk of veering into pseudo-science, I would like to point out that most of us have rather strange cognitive experiences in our lives, such as deja-vu (which may be precognitive. I have had such precognitive deja-vu episodes), or empathy/telepathy/etc.
Discussing these topics in serious scientific circles will get you funny looks, but each one of us knows that we have experienced them, these experiences do play roles in our lives.
It occurs to me that some of the freaky characteristics of quantum could account for these experiences if, like birds, we have some binding between perceptrons of some kind and quantum coherence effects. What we perceive as psi may simply be quantum perceptions. That certainly would entail survival advantages, wouldn't it?
Ako: I have written some previous blog posts on psi, and I tend to believe it is a real phenomenon. However, quantum theory in itself doesn't seem to convincingly explain psi phenomena, though it may be part of the story. Check out Damien Broderick's book "Outside the Gates of Science"....
Obat kutil kemaluan // Obat kutil kemaluan // Obat kutil kemaluan // Obat kutil kemaluan // Obat kutil kemaluan
Post a Comment