I'm interested in whether the scale of processes that occur in the brain is small enough to be affected by quantum mechanics. For instance, we ignore quantum mechanics when we analyze a game of tennis because a tennis ball is much too large to be affected by quantum mechanics. However, signals in the brain are mostly (all?) electrical, carried by electrons, and electrons are definitely 'small' enough to be affected by quantum mechanics. Does that mean the only way we will be able to further understand how the mind works is through an application of quantum mechanics?
$\begingroup$
$\endgroup$
18
asked Jan 10, 2016 at 9:08
-
6$\begingroup$ Mammalian brains and CNS use completely classical electrochemical processes. The molecular isomerization timescales for those processes for which quantum mechanics is vital, like in the signal reception in the eye, are on the order of ps while neural functions are happening at the ms timescale... so, no, there is no quantum mechanics involved, as far as one can rationally tell by comparing molecular processes with information processing. $\endgroup$– CuriousOneCommented Jan 10, 2016 at 9:23
-
25$\begingroup$ @buzhidao: That's not physics but physicists with some very questionable philosophical ideas making money by booksales. $\endgroup$– CuriousOneCommented Jan 10, 2016 at 9:25
-
8$\begingroup$ I'm of the camp that this question is still not answerable with our modern technology, we need more empirical investigations; theories be damned. $\endgroup$– LoonuhCommented Jan 10, 2016 at 18:23
-
38$\begingroup$ Quantum mechanics plays a major role in my brain, because thinking about it is my job. $\endgroup$– rob ♦Commented Jan 10, 2016 at 21:37
-
9$\begingroup$ Quantum mechanics has a role in everything in the universe. What role it has, and whether it's worthwhile to consider this role, is a different matter. $\endgroup$– SuperbestCommented Jan 10, 2016 at 23:02
|
Show 13 more comments
8 Answers
$\begingroup$
$\endgroup$
29
Quantum mechanics has almost no bearing on the operation of the brain, except insofar as it explains the existence of matter. You say that signals are carried by electrons, but this is very imprecise. Rather, they are carried by various kinds of chemical signals, including ions. Those signals are released into a warm environment that they interact with over a very short timescale.
Quantum mechanical processes like interference and entanglement only continue to show effects that differ from classical physics when the relevant information does not leak into the environment. This issue has been explained the context of the brain by Max Tegmark in The importance of quantum decoherence in brain processes. In the brain, the leaking of information should take place over a time of the order $10^{-13}-10^{-20}$s. The timescale over which neurons fire etc. is $0.001-0.1$s. So your thoughts are not quantum computations or anything like that. The brain is a classical computer.
-
20$\begingroup$ "The brain is a classical computer" - I don't think that conclusion is necessarily supported. Certainly the brain is very bad at doing things that are trivial for a classical computer, but quite effective any many things that are extremely difficult for a computer to replicate. It may not be analogous to a quantum computer, and it may be a purely deterministic construct, but that doesn't mean it therefore fits the mold of a classical computer/can be accurately modeled by a deterministic Turning machine. $\endgroup$– arothCommented Jan 11, 2016 at 4:29
-
15$\begingroup$ @aroth: I note that "extremely difficult in principle for any computer to replicate" and "we don't know how to build a computer that replicates this -- yet" are very different things. I have a computer in my pocket that I can ask in English when my wife's birthday is, and it can speak the answer to me. That was absurdly impossible not even fifty years ago. $\endgroup$ Commented Jan 11, 2016 at 4:44
-
10$\begingroup$ @EricLippert And would have been unequivocally interpreted as a sign of intelligence not even 50 years before that ( pre-Turing). (Following Ray Kurzweil's observation that anything accomplished by machines is ex post excluded from the definition of intelligent.) $\endgroup$ Commented Jan 11, 2016 at 4:53
-
9$\begingroup$ @EricLippert - Certainly (though I'd say my anecdotal experience with voice recognition software is significantly less polished than what you describe). But if making a factual assertion that "the brain is a classical computer", there should be some proof to back that up. To the best of my knowledge it hasn't been demonstrated, either mathematically or by way of practical example, that the brain is definitely a classical computer (or any sort of computer, as traditionally defined within the realm of computer science). $\endgroup$– arothCommented Jan 11, 2016 at 5:44
-
9$\begingroup$ @aroth I think it is fairly obvious that a brain is turing complete. It is possible for a human controlled by a brain to carry out all the functions of a turing machine. By extension, we can do anything that a computer can do given sufficient time and storage/memory (e.g. pen + paper). The fact that the brain is very bad at it is irrelevant. Computers from 30 years ago were very bad at things that are trivial for today's computers, but we still consider them computers. $\endgroup$– JBentleyCommented Jan 11, 2016 at 10:24
$\begingroup$
$\endgroup$
1
Yes - but only in the sense that all macroscopic processes depend on underlying quantum mechanics at the microscopic scale.
No - quantum mechanics is not the best model for describing what happens in the brain.
In one sense, the behaviour of a neuron is similar to a quantum process, such as (for example) the decay of an electrically excited or radioactive atom to its ground state. A neuron either fires or it doesn't. But there are many machines that either fire or not, so this is not sufficient for us to infer that this a quantum process.
https://en.wikipedia.org/wiki/All-or-none_law
There are some important differences as follows (the most important of which is the scale of the process.)
The atom emits a photon (a single quantum of electromagnetic radiation) randomly, and independently of events in its environment (for spontaneous emission, at least). We can know experimentally what the probability of a particular type of atom emitting a photon in a particular time period is.
The neuron emits a an impulse (a large number of ions) in a fairly predictable manner, depending on the impulses and stimuli it has received. A good (if rather basic) model of this would be a water tank that automatically empties itself when it is completely full. Such tanks are used to flush urinals in mens' public bathrooms. It's a big step from this to building a computer as sophisticated as the brain, but it should be clear that such a tank does not depend on quantum mechanics. Note that an electronic analogue of such a tank is possible.
-
$\begingroup$ How I read your explanation, a neuron fires or not, suggest a lack of specificity in your model in the concept of memory. That's why I would argue to a deeper structure that has patterns to convey different information. Also, recruitment of the particular neurons and their spindle fibers has been postulated to use Quantum Tunneling, $\endgroup$ Commented Oct 28, 2016 at 11:16
$\begingroup$
$\endgroup$
4
Roger Penrose and Stuart Hammeroff are working on this exact hypothesis. They believe the spindle fiber is the structure that collapses the Quantum Wave Function. As of now, they have been unsuccessful in showing the "spindle fiber" in supporting QM abilities. But there is plenty of info on the subject, starting with Roger Penrose book "The Emperor's New Mind".
-
12$\begingroup$ Penrose is a perfect example of a mathematical genius who has, at least of lately, trouble screwing in a light bulb. No offense... but these questions are experimentally decidable and Penrose has, IMHO, done absolutely nothing to follow the exiting experimental evidence. $\endgroup$ Commented Jan 10, 2016 at 16:37
-
$\begingroup$ If they think there is a collapse of a wave function needed, and that they've located the magical device that does it, then indeed they don't understand even basic quantum mechanics. $\endgroup$– user10851Commented Jan 10, 2016 at 20:06
-
3$\begingroup$ My reading of "The Emperor's New Mind" is that they ascribe brain function not only to quantum mechanics, but to quantum gravity. As the density of matter in the brain is not very different from water, this makes even less sense to me. $\endgroup$ Commented Jan 11, 2016 at 0:36
-
$\begingroup$ @Ross Millikan - Penrose's idea was that there is some threshold of difference in the metrics, beyond which a superposition of space-time geometries is not well-defined, and that that is when "wavefunction collapse" occurs. And the threshold is supposed to be quite subtle, so that it can come into play on mesoscopic scales. $\endgroup$ Commented Jan 11, 2016 at 2:24
$\begingroup$
$\endgroup$
11
The brain is a de-facto a classical computer as explained in Alanf's answer. However, this then leaves open the possibility that what makes a classical system like our brain or some future AI conscious, could well be related to how quantum mechanics reduces to classical mechanics. One proposal along these lines (that I personally don't find compelling) has been put forward by Roger Penrose as mentioned in Ed Yablecki's answer.
A far simpler idea is to consider that quantum mechanics in the classical regime is still not the same as classical mechanics. What happens is that due to entanglement with a huge number of environmental degrees of freedom, many typical quantum effects are effectively lost and you can then pretend that they don't exist. As far as predicting the outcome of experiments is concerned, you can use classical mechanics with impunity. But the physical system is not what you get when you take its classical description to be literally correct.
You can see clearly how the difference between the exact quantum mechanical description of an AI + environment and the classical description of the AI answers a lot of the philosophical objections against the strong AI hypothesis. In the exact description there is plenty of room to invoke correlations between inputs and outputs as existing at any particular moment, because what the AI experiences is only a coarse grained measurement that is consistent with a large number of microstates. These then exist as parallel worlds within its de-facto measurement error. The real existence of such an ensemble of correlated states defines which computation is actually being performed at any given instant. The difficulty of doing that within a purely classical picture is at the heart of the criticism of strong AI.
Consider Marvin Minsky's famous thought experiment of simulating your brain by a huge analogue device consisting of huge wheels and gears. Then strong AI says that this simulation will succeed, but the critics say this is just ridiculous, how on Earth can a collection of wheels and gears feel anything at all? They key observation to be made i.m.o. is as follows. From your point of view, the exact state of the wheels and gears cannot be pinned down precisely. While you can look down on several of your wheels, any attempt by you to find out the state of all your wheels will fail due to your memory having a finite capacity; most of that capacity is used to run the programs that define you. Whatever you feel, whatever consciousness really is, it is ultimately a computation and a system of wheels and gears can define that unambiguously, provided you invoke the MWI.
answered Jan 10, 2016 at 17:27
-
1$\begingroup$ You are so wrong. The brain is NOT a classical computer at all. The simple prove is that humans can decide the halting problem, the Turing machine cannot. A computer cannot compute recursively-non-computable languages. $\endgroup$– JerrynoCommented Jan 12, 2016 at 13:23
-
9$\begingroup$ @Jerryno Humans can't do that either! Humans can certainly find some halting problems - but then so can Turing machines. The point is that there is no way to 100% identify all halting problems, and that applies just as much to humans doing the identification. $\endgroup$– GrahamCommented Jan 12, 2016 at 14:15
-
$\begingroup$ @Graham true, but that is because the brain is not deterministic so you cannot build a prove around it. Humans can decide some halting problems that Turing machines can't. If a brain was a classical computer you could foretold everything about it given the initial conditions. Which is clearly not how it works. You should watch this to the end: youtube.com/watch?v=sMb00lz-IfE everything is explained very nicely there. $\endgroup$– JerrynoCommented Jan 12, 2016 at 14:22
-
$\begingroup$ @Jerryno Untrue - humans cannot do that. The only way in which a human can do that is if they were given information which the Turing machine did not have. Give the Turing machine the same information, and it can solve it too. $\endgroup$– GrahamCommented Jan 21, 2016 at 17:15
-
1$\begingroup$ @Jerryno: what you are describing is essentially a probabilistic Turing machine. Such machines are no more powerful, in the sense of decidability, than any other TM, because they are a special case of an NDTM (non-deterministic TM). Whether they are more powerful in the sense of complexity (can they solve problems in polynomial time which a DTM can't?) is, I think, not known but thought not to be the case. $\endgroup$– user107153Commented Mar 9, 2016 at 11:22
$\begingroup$
$\endgroup$
In fact electric charge is carried by positive ions (sodium and potassium), not electrons, along the neuron's axon. Are they small enough? I don't really know. By the way, there is a hypothesis that a certain species of birds use a pair of entangled electrons to orientate themselves. Quantum mechanics also plays a role in enzymes (quantum tunneling for example) but due to the large amount of enzymes I don't think quantum effects make a difference. There is a good video that shows some hypothesis of how some organisms make use of quantum mechanics (it's not specific to the brain, though).
As for the brain itself the answer is no one knows, maybe it's used to store information or who knows what. As you said it'd be a good thing to take quantum mechanics into account (sometimes at least) to try to understand more of the brain.
$\begingroup$
$\endgroup$
There are definitely some serious scientific efforts going on today trying to explain and incorporate QM in brain processes. The following TED talk is just about that:
Other scientists hypothesize that the tiny dendrites or microtubules on neurons inside brain are the interface where the feeble quantum effects take place and create the known classical effects.
The Physician Mark Germaine writes in his essay: “A wealth of data supports the notion that the dendritic arborizations are the primary structures that support perception (Pribram, 1991). The neural wave form characterizes the dynamics of the dendritic network, and this wave form can be described by an equation that is fundamentally the same as the equation describing the quantum wave form (Pribram, 1991),………”
But he is discussing other model and not directly the QM model for brain. Find the essay here:
http://dynapsyc.org/2015/HOLOMIND.pdf
The point is that the commonly accepted mechanistic model for brain does not look as being enough to account for the entire richness of experience – not even the seemingly most basic ones like taste, smell, vision. The mechanistic model cannot account for the enormous versatility of any consciousness – not even as simple as of an ant.
$\begingroup$
$\endgroup$
1
I'm late to the party but didn't see the following argument, which is at the intersection of ethics, physics and logic. It has to do with the free will which we perceive we have.
The logical definition of free will is that it is not determined; that's the meaning of free. We do something not based on rules, i.e. we do it not predictably, but spontaneously. We could as well have decided differently, but we didn't.
In general, we do not see predictable behavior as signs of free will: Such behavior is guided by instinct, morale, conventions or external influences like advertisements or peer pressure.
Signs of free will, by contrast, are visible at cross roads: Somebody decides to be a hero or not, eat at one restaurant or another. We, and possibly the person herself, didn't know beforehand.
Generally spoken, all behavior falls in one of these two categories: It is rule based, i.e. predictable; or it is a decision out of our free will, which nobody could reliably predict.
The interesting thing is that from a computational standpoint, "unpredictable" is simply equivalent to "random". That's the definition of "random": It does not depend on previous events, i.e. there are no rules to predict a random event from previous events.
Free will is logically the ability to make random decisions.
(As an aside, this is the reason that there will be no categorical obstacle to emulating human behavior with computers. All behavior is either rule-based or random. We can emulate rule-based behavior very well with computers. But it's also not that hard to introduce or emulate randomness.)
Let's not fool ourselves. Much of what we perceive as free will is not free at all; the overwhelming majority of our behavior is governed by our culture, ethics, taste, principles and so on, often unconsciously. That we seem less predictable than we actually are is owed to a lack of information on the side of the observer (and big data is telling us that we do become fairly predictable -- and gullible -- given enough information about us).
But we do have a strong feeling -- and the occasional examples -- of personal freedom. If we were mechanical, deterministic clockworks, we wouldn't be free. But we are not clockworks; the working of our brain is not entirely deterministic.
The physical underpinning of this indeterminacy must be random events in our brain. Electrons, atoms and molecules are not billiard balls; no Laplace's demon could predict a brain's future, principally not. The microscopic world which is the underpinning of the world we perceive is simply not deterministic. In a non-linear system as the brain small quantum events which could have gone either way will occasionally make a difference. A neuron fires, or not; a group of neurons' excitation just crosses the threshold in a contest with another group to become dominant, or not.
In this sense, as a source of innate indeterminacy, and thus freedom, I believe quantum effects play a huge role in our brain. And in the universe.
answered Mar 9, 2016 at 11:00
-
$\begingroup$ A good point Peter A when you say that given enough info the overwhelming number of are choices are predictable. My free will does not choose the resturant but the sum of all my preferences, experiences and present state of mind makes the choice unconsciously. Nobody has a free will - they just don't know it! $\endgroup$– JensCommented Mar 9, 2016 at 15:58
$\begingroup$
$\endgroup$
Brain as a biocomputer is far to complex to expect an fully descriptive answer in a form of a forum post.
Don't worry, the situation is even more riddled! :) The opinions on the topic of quantum behaviour affecting the perception of reality (and accordingly creation of "reality", but I am touching very metaphysical viewpoint here, which is usually avoided here, and for good reasons) vastly differ.
All in all, in my opinion the currently correct answer is "Due to complexity of the brain, we do not know, where the quantum phenomena apply and where not in terms of brain and perception." At this point, one has to join the research to start getting at least partial answers.
"(...) I don't feel frightened not knowing things, by being lost in a mysterious universe without any purpose, which is the way it really is as far as I can tell." R. P. Feynmann
