Here’s a 20-minute segment film of an episode of the BBC documentary series Horizon called Humans v2.0, featuring Ray Kurzweil:
[See the whole episode at Google Video]
About one-and-a-half minutes into the clip, the narrator says that “he [Kurzweil] believes that our understanding of the human brain will soon be complete.”
Although this is not a direct quote from Kurzweil himself, he does actually believe this. It is true that advances in our understanding of the brain are coming thick and fast. But I’m certain pretty sure that we’ll never fully understand the brain, and think that Kurzweil’s claim is utterly ridiculous.
Why do you think the brain will never be fully understood? It seems like a strange scientific viewpoint.
It seems equally strange that you think all neuroscientists should believe that the brain can be fully understood.
That’s being very cryptic again 🙂 What about the brain do you think will never be fully understood? Just curious. If there is some aspect of the brain that is not amenable to scientific comprehension, I would think that is a strange scientific view. But I am not sure what you mean-
What I mean is this: just because someone is a scientist does not necessarily mean that he or she believes that science can provide all the answers. In my case, I certainly do NOT think that science can provide us with a complete understanding of the human brain. And I am not trying to invoke a teleological argument here; my point is that the complexity of the brain is unfathomable, and cmoplete comprehension of it is, in my opinion, impossible. Yes, we can, and do, make discoveries about many aspects of brain function. But the more we learn, the more complex the brain becomes, and the more confused we seem to get.
But I may be wrong. Only time will tell.
My biggest problem with a lot of what Kurzweil says is that he completely underestimates the complexity of human biology. We’ve only scratched the surface. There is a long long way to go. Will we get there. Eventually I think we will, but not in Kurzweil’s lifetime
May we be discussing over semantics. Few things in science will be “fully understood” since, basically, that would spell the end of science. And it may be tough to conceptualize how a mirror can “fully understand” itself.
On the other hand, despite the complexity of the brain and the natural confusion that “new” science branches go through, I don’t see why “understanding of the human brain” should be that different from understanding in other scientific disciplines.
I must agree with A.G. that it is somewhat strange for a scientist to state, that we will never understand the brain. What is it exactly about the brain that makes the understanding of it absolutely impossible. You say that “the complexity of the brain is unfathomable”, but why do you think so? I guess the brain is more complex than say the liver or the heart, but the difference is a quantitative (i.e. not a qualitative) one, so – given that we can fully understand the liver and the heart – you are postulating some upper level of complexity limiting our cognitive abilities. Why do you think there is such an upper limit and how should it be possible to determine it (don’t tell me that you think the brain is unable to understand something of the same or higher complexity than it self)?
To be honest, your claim reminds me somewhat of the IDiotic claim that the complexity of living structures makes it unfathomable that they could have evolved by natural processes and that they must therefore have been designed.
Morten – you guess the brain is more complicated than the liver or heart? Well I guess they’ve never made a fully functional artificial brain.
I take it you’re talking about what Intelligent Design creationists call irreducible complexity. And you’re equating my use of the word “unfathomable” with the meaning of “irreducible” when used by creationists. I reduce the brain to nuts and bolts every day on this weblog and talk evolution every other day. So if you’re accusing me of being a creationist, you’ve got your wires crossed.
Of course we can gain a rough understanding of the brain from gross anatomical to the subatomic level. But know every fine detail of every level AND integrate all the levels coherently? Or fully understand the dynamics of neural development? Or consiousness? I don’t think so. Not in this century anyway, and probably not the next. Maybe never.
Alvaro’s right. It’s all about semantics. We have a different understanding of what “complete understanding” and “unfathomable” actually mean.
MC, I would never dream of accusing you of being a creationist – I read your blog almost every day (and like it very much by the way). I was merely pointing out, that I find the reasoning in your claim that the brain is so complex that we will never understand it analogous to the reasoning in the IDist claim that life is so complex that it could not have evolved. Both claims are unsubstantiated and not supported by any evidence and seems to be simply based on a gut feeling. If you disagree, you should come up with more substantial arguments than just “I don’t think so”.
You and Alvaro may be right, that it’s mostly a question of semantics. However, I would argue that if a complete understanding is only achieved when we “know every fine detail of every level AND integrate all the levels coherently”, then we will never have a complete understanding of anything in biology (or meteorology or cosmology or …) and then you should say that in stead of singling out the brain. This was actually what I meant, when I said “I guess” the brain is more complex than the liver or the heart. You will e.g. never know the position and velocity of every atom or the spin of every electron in the heart, but wouldn’t you agree that a complete understanding of the heart as a biological organ can be achieved without this knowledge? I certainly do, and likewise I think that a “complete understanding” of the brain as a biological consciousness generating organ could be obtained without knowing what role every single AP or every single release of a neurotransmitter vesicle in every single synapse is playing.
I think that to answer the question whether the brain can ever be “fully” understood someone should define the meaning of “fully” in practice, in wet biology. Who would need to understand the movement of each subatomic particle in the brain in each moment under all possible circumstances (that is the closest to the “full” understanding that I can imagine)? My objection to the Kurzweils hyperoptimistic predicition is that we should not dare to talk even about the sufficient understanding of the brain before we know enough to be able to substantially help schoziphrenics, autistics, Alzheimer patients, stroke victims etc. These are the practical needs, everything else is a wasting of energy on academic questions.
I think we are all saying the same but from different perspectives.
If I understand MC’s main criticism, it is about timing of when we will have a “substantially useful understanding” (let’s forget “full”!), which is his view is way farther away than what Kurzweil says. To make the debate more specific and constructive, maybe we should propose what 5-10 applications of that knowledge we could expect by when, and then contrast our projections, which will help crystallize our assumptions.
MC: may I ask what, in your view, will the 2-3 key practical contributions of neuroscience to society over the next 5 years be?
I cannot, of course, be 100% certain that we’ll never fully understand the brain. Perhaps “certain” is too strong a word, and Kurzweil’s claim is very, very over-optimistic rather than utterly ridiculous.
Don’t talk like that man. It makes alarm bells go off in my head left and right.
At least Kurzweil has a case to back up his claim.
The brain, and every other system, is infinitely complex. We will never completely understand anything, and I’m consistantly stunned that so many members of the scientific community think otherwise. How long have we been claiming that full understanding of this or that is just on the horizon? Since as long as we have recorded history.
It makes me sick to my stomach reading phrases in academic textbooks like “only recently have we come to fully understand the workings of the universe/body/environment”. Bullshit. How can this be such a prevalent flaw in our thinking?
At last! Someone who seems to fully agree with me!
Right on brother.
Re: “How can this be such a prevalent flaw in our thinking?”
Because “understanding” is a moving, mutable target. On one hand, we clearly understand things that we previously had not, and there is no reason to believe that this process will not continue. But on the other, there is always a way for one to insist that we don’t really or completely understand a given thing.
This is reminiscent of early AI debates, where people would say things like “well, a computer won’t be able to X, because that requires intelligence”, where X is something like beating a human at chess. When the computer beats the humans, then people say “well, the computer did X in a non-intelligent way” or “well, we were wrong, and doing X doesn’t really require intelligence”, and X is changed into something else.
In the same way, any particular criteria for “understanding” can be given, and then researchers meet those criteria, and say things like the quotation in the hypothetical textbook. But then, the target for “understanding” gets moved again, and we’re back where we started.
My two cents, anyway.
I don’t fully understand this discussion
I think part of the issue is semantics… “fully” understood is a moving target. As an engineer, I’m reminded of the old joke about the mathematician and the engineer. If one was located on the other side of the room from a lovely woman, and one was allowed to take steps equal to half the distance toward her each time, the male mathematician would say, “That’s useless, I’ll never get there.” The male engineer would say, “Yes, but you’ll get close enough…” Similarly, while we may never fully understand the brain, it seems possible that we’ll understand enough to explain various behaviors and other phenomena.
One of Kurzweil’s predictions that I’m watching is that resolution of brain scans will improve exponentially. While the resolutions improve in fits and starts, so far I’d say he’s on track.
Justin, of course we will never completely understand anything in the absolute sense – we cannot even be absolutely sure that the physical, material world actually exists. However, this fact is not really interesting or of any real significance. It does not mean that we should halt our scientific efforts to gain an increasingly better understanding of the world and it certainly does not mean that alternative ways of “knowing” should be considered equal to the scientific way (as postmodernists and relativist might argue).
Furthermore, it has no bearing on the question of understanding the brain (as opposed to understanding e.g. the heart). We have a very good (satisfactory?) understanding of the heart although this understanding does not include knowledge about what every elementary particle in the heart is doing (actually all this information would be irrelevant). I see no justified reason to believe that we should not be able to gain a similarly satisfactory understanding of the brain in the hopefully not to distant future.
I’m sorry if this makes you sick to your stomach but maybe you should concentrate on producing some evidence or good arguments in stead of obsessing about your gut feelings.
I don’t think there is much utility in qualifying understanding in terms of its completeness without predicating an ends or a purpose to the understanding. With five to seven thousand different types of neurotransmitters, thousands of possible connections between any two neurons, and other statistics of that staggering scale, I’d say that it’s bit obtuse to claim that it would ever be possible to achieve a complete understanding of the brain. But that isn’t just because of the complexity of the subject; the idea of the complete understanding of anything gets on my nerves.
Thanks for the thought-provoking post and for a great blog.
Curtis – according to the textbooks, a single cerebellar Purkinje cell forms synapses with about 100,000 parallel fibres. I’m willing to stick my neck out and say that it forms 1,000,000 synapses. Either way, the complexity of the brain is staggering, and I stand by my claim that we’ll never fully understand it.
A glass of water contains maybe 5000000000000000000000000 water molecules that move around and bump into each other all the time. Does that prevent us from understanding a glass of water? A living cell contains maybe 10000000000000 molecules of perhaps 10000-20000 different kinds. Does that prevent us from understandig the living cell? From actually understanding what life is (something that seemed utterly impossible 100-150 years ago)?
Granted, the brain is more complex than a glass of water or a single living cell, but where is the evidence that its complexity is above some upper limit of human understanding (and that such a limit indeed exists)?
I find it very strange to apriori rule out our ability to understand the brain or anything else for that matter. Why do some of you feel this apparently strong need to do so? Why not just wait and see?
The water-molecules analogy is just kind of silly. You can make observations on the interactions between the molecules that generalize across the whole glass, so the difference in interactions between any two bunches of molecules are essentially irrelevant to your understanding of the glass. A brain consists of all these trillions of connections doing importantly different work for the purposes of understanding mental goings-on.
Even assuming (perhaps reasonably) that our machines will attain the computational capacity of a brain, it’s very likely that there will be physical constraints preventing us from ever observing all the activity we need to “fully understand” the processing. If every neuron in some nucleus is doing something important but we can’t get the spatial or temporal resolution record the activity of each one simultaneously and quickly, the depth of understanding have a hard-and-fast limit.
Yonatron, you can also make observations on the interactions between neurons that generalize across the whole brain. The point of the water analogy simply is that it is not necessary to know what every single molecule is doing and exactly where it is in order to understand water. Similarly, I find it quite likely that it will be possible to understand the brain without knowing the exact firing pattern of every single neuron in the whole brain (I guess I agree that it might never be possible to achieve the resolution needed to obtain this information). First of all, the strong dependence on the activity of every single neuron that you suggest is unlikely because it would make the system extremely vulnerable. Secondly, many neurons are working in parallel doing essentially the same thing. Thirdly the brain is modularized, which makes it easier to analyse.
Note that I am not saying that we will fully understand the brain in 10, 20 or 100 years or even that we ever will. I’m just wondering why you people insist on apriori ruling out that we will . Your view is not supported by any evidence, but seems to be fuelled by strong emotions. Are you very emotionally attached to the idea of “the mystery of the human brain” or what is going on?
By the way, do you also reject the possibility that we will be able to understand the dog brain, the crow brain, the frog brain, the bee brain …. Or is it just the human brain?
Actually Steven Pinker has explained (to Stephen Colbert) how the brain works in just five words: Brain cells fire in patterns.
See video at: http://www.comedycentral.com/motherload/index.jhtml?ml_video=81914
Morten, I didn’t go so far as to reject the possibility of understanding the human brain–I actually just returned to college many years after dropping out so I could get a bachelor’s in cognitive science. I think “mysterians” are irritating and I hope to contribute to our knowledge of the brain someday. I just think that Kurzweil and people like him make pronouncements with insufficient caution. They reject the possibility of permanent roadblocks to understanding. I don’t know for sure that any such roadblocks exist, but they seem plausible.
We probably agree more than we disagree, but I still think the water-molecules analogy glosses over differences in complexity. We may not need to know about all the trillions of synapses in a brain to “get” what it’s doing; but even if chunk them together and look at interactions between huge subsets of connections, that’s still vastly more information than what’s relevant to a uniform glass of water.
Yonatron, I’m glad that you dislike mysterians – we certainly agree on that. One of the reasons that I keep insisting that it is not a good idea for scientists to announce roadblocks to our understanding of the brain is exactly that it provides ammunition for the mysterians. All kinds of quacks (religionists included) will just love to state that “neuroscientists say that we will never understand the brain” and use that to promote their own whacky supernatural beliefs.
The brain is not an infinitely complex organ and it will eventually be understood, but not soon. You have to realize that the NIH is the economic engine behind neuroscience research, but their whole funding system is disease-oriented, not fundamental science oriented. Most neuroscientists are studying little, isolated parts of the system: molecules and cells. This is ultimately driven by the search for effective pharmaceuticals. Very few neuroscientists study anything that has to do with information per se, let alone whole integrative theories of how brains work as coherent informational systems. The little modeling that is done is focused on the behavior of single neurons and local circuits. The barrier is that we do not understand the nature of the signals in the wires, i.e. the neural codes that subserve representations for sensory inputs, internal models, and action, let alone the means by which such encoded information is stored in long-term memory and later retrieved for use. Without some understanding of the nature of the signals, the biophysical models are next to useless in getting to a theory of the brain as an informational engine that generates its own meanings in pursuit of its own purposes.
Ray is just naively overoptimistic about how well brains are presently understood. I wish he were right. He has been reading the hyped press that scientists generate in order to justify the existence of their own research programs. When the chances of getting your NIH grant funded are well below 10%, as they are now, scientists become very reticent to admit of any fundamental limitiations of or difficulties with their approaches. Understanding brains will happen once enough people are allowed to put their minds to the problem in a manner that is focused on the basic scientific questions. Guaranteed to happen, but maybe not in our lifetimes.
Well, it all seems a tad bit premature to talk of understanding the brain, when even the foundations of neuroscience are still theoretical – I mean the Neuron doctrine. If someone wishes to make such claims, they are very optimistic…
For 30 years I have actively programmed Slater’s PCA based Ingrid in line with Personal Construct Theory. I also push around my own ideas to do with the things Peter talks about above, namely the understanding of “the nature of the signals in the wires” and how “encoded information is stored in long-term memory and later retrieved for use”. I continuously develop these concepts as an electronic cocoon which my brain constantly interacts with. Hardware hopefully will be available for me to program brain-only life support metallization by the time I need to separate from my body. That said, there is still the possibility that the brain’s information structures may not be all that is needed to reconstruct the memories that serve the mind. As such, I allow for a weird thought experiment that says the brain’s neural nets may only harness the neural matter in question which in turn then provides wormholes back to the reinforcing events in solid streams of space-time. Not only can I understand the implications that the brain might be an ‘other worldly’ gateway to other such realms of reality, I can also conceive of creating a totally digital embodiment that can still act as an interface to sufficient robotics to always consider myself to be alive. So while you feel that you may know the irony and futility of my pursuit, I’m having too much fun to be bothered needing help from the likes of the NIH.