Creating vs treating a brain

The NAND (Not AND) gate is all you need to build a universal computer. In other words, any computation that can be done by your desktop computer, can be accomplished by some combination of NAND gates. If you believe the brain is computable (i.e. can be simulated by a computer) then in principle, this is all you need to construct a brain. There are multiple ways to build a NAND gate out of neuro-wetware. A simple example takes just two neurons. A single neuron can act as an AND gate by having a spiking threshold high enough such that two simultaneous synaptic events are required for it to fire. This neuron then inhibits the second neuron that is always active except when the first neuron receives two simultaneous inputs and fires. A network of these NAND circuits can do any computation a brain can do.  In this sense, we already have all the elementary components necessary to construct a brain. What we do not know is how to put these circuits together. We do not know how to do this by hand nor with a learning rule so that a network of neurons could wire itself. However, it could be that the currently known neural plasticity mechanisms like spike-timing dependent plasticity are sufficient to create a functioning brain. Such a brain may be very different from our brains but it would be a brain nonetheless.

The fact that there are an infinite number of ways to creating a NAND gate out of neuro-wetware implies that there are an infinite number of ways of creating a brain. You could take two neural networks with the same set of neurons and learning rules, expose them to the same set of stimuli and end up with completely different brains. They could have the same capabilities but be wired differently. The brain could be highly sensitive to initial conditions and noise so any minor perturbation would lead to an exponential divergence in outcomes. There might be some regularities (like scaling laws) in the connections that could be deduced but the exact connections would be different. If this were true then the connections would be everything and nothing. They would be so intricately correlated that only if taken together would they make sense. Knowing some of the connections would be useless. The real brain is probably not this extreme since we can sustain severe injuries to the brain and still function. However, the total number of hard-wired conserved connections cannot exceed the number of bits in the genome. The other connections (which is almost all of them) are either learned or are random. We do not know which is which.

To clarify my position on the Hopfield Hypothesis, I think we may already know enough to create a brain but we do not know enough to understand our brain. This distinction is crucial.  What my lab has been interested in lately is to understand and discover new treatments for cognitive disorders like Autism (e.g. see here). This implies that we need to know how perturbations at the cellular and molecular levels affect the behavioural level.  This is an obviously daunting task. Our hypothesis is that the bridge between these two extremes is the canonical cortical circuit consisting of recurrent excitation and lateral inhibition. We and others have shown that such a simple circuit can explain the neural firing dynamics in diverse tasks such as working memory and binocular rivalry (e.g. see here). The hope is that we can connect the genetic and molecular perturbations to the circuit dynamics and then connect the circuit dynamics to behavior. In this sense, we can circumvent the really hard problem of how the canonical circuits are connected to each other. This may not lead to a complete understanding of the brain or the ability to treat all disorders but it may give insights into how genes and medication act on cognitive function.

Advertisements

8 thoughts on “Creating vs treating a brain

  1. What is the physical scale length of the descriptive canonical circuits that you are proposing? micrometer? millimeter?

    Like

  2. @Tom
    Great question and I don’t really know. The concept is more based on function than size but a reasonable working hypothesis is that it is on the order of a hundred microns. In any case, it is certainly larger than a micron and perhaps less than tens of millimetres. There is likely to be no fixed size and the effective size of a circuit may even change depending on task.

    Like

  3. Hi Carson,

    (Just for the sake of discussion) you mention that you hope to explain autism with abnormalities of your simple canonical cortical circuit, because this simple circuit can explain working memory and binocular rivalry. (Or, at least I think that you are saying that.)

    So, if this group found that people with autism have normal binocular rivalry (http://www.journalofvision.org/content/12/9/1365.short) which would suggest a typically functioning canonical cortical circuit, then would that force you to reconsider that perhaps the fundamental defect in autism may be something other than an abnormal simple canonical cortical circuit?

    Tom

    Like

  4. Hi Carson,

    In fact, I do believe the brain is computable, but I think you agree that the Hopfield Hypothesis must go beyond that statement to be a meaningful new hypothesis.

    When you wrote earlier, in your post from 2006, “that perhaps we already know all the neurophysiology we need to understand the brain but just haven’t put the pieces together in the right way yet”, I guess, you did not merely state there that the brain is computable, right? As you saw from my recent comment, I focused on the problem how-to put pieces together. Why?

    Because: I don’t think we should or even can separate in a meaningful way neurophysiology (function) from anatomy or “putting pieces together” (structure).

    But let me go one and formulate two other hypotheses.

    1. The brain cannot be build in a single progressive assembly line fashion, that is, either evolving like in biological evolution (phylogeny) or like in embryonic development (ontogeny). A artificial brain needs both these lines, phylogeny and ontogeny, to build the proper circuits that can do any computation a natural brain can do.

    2. The artificial brain cannot be build in its final stage even if firmly committed to (1) only with one interface to its environment.

    Earlier, I argued that without any closed-look interface, it is not a brain, so (2) is stronger. How many separate sensory systems do we need (or will have to develop during the process)? And is pain part of it? I do not know. But for the sort of organized complexity we are looking at, a single one is not sufficient (I believe). Personally, I think two are sufficient one of which is pain.

    Can we change the order of phylogeny and ontogeny in some way? So that these distinct processes do not run like in our brain’s development but alternate and run in parallel in some new fashion? Again, I do not know, but I certainly think that there are several routes to the final brain. But infinite many routes?

    While there is an infinite number of ways to create an NAND gate out of neuro-wetware and this implies that there are an infinite number of ways of “set up” a brain, this does not imply that there are also the same number of ways to build (develop) it over time. Of course, by the hand of an almighty creator an infinite number of brains could exist, but we talk about developmental rules, right. Creation from a few design principles, because we have the treatment in mind that probably depends one this.

    These two hypotheses, which could be completely wrong of course, interlink physiology and anatomy. To my mind, we currently lack fundamental knowledge how exactly they are linked.

    In this sense, I’m not a proponent of the Hopfield Hypothesis.

    So, we have also to be careful with respect to treat the brain. By the way, will you be at the MBI for the Workshop 3: Disease?

    http://mbi.osu.edu/2012/ws3description.html

    Like

  5. @Tom I Yes, I know of this paper well. In fact, we’re working on a theoretical paper showing that depending on how you perturb the cortical circuit any number of things can happen including nothing. Also, keep your eye out for some new data in the next year or so that will mix things up a bit.

    @Markus It depends on what you mean by creating a brain. What I mean is to create an artificial intelligence that can replace any job a human currently does. I believe it is just a matter of time for us to do that and we don’t need to know any new biology. As I capitulated in my post, it is very likely we do not know enough to understand our brain or biological brains in general.

    I won’t be going to the MBI Workshop. (I didn’t get invited:)).

    Like

  6. Hi Carson,

    yes, but now it depends on what you mean by “any job a human currently does”.

    Certain jobs, which dependent upon creativity, have only be done by human beings suffering from mental disorders or neurological diseases. Picasso’s cubist style of painting may have been inspired by a cubist visual migraine aura, likewise Giorgio de Chirico’s unrealistically deep perspective. Let’s replace job by behavior and if we then also acknowledge that the division into healthy and pathological behavior is man made, we’re getting somewhere. (Actually this is interesting: If this job — defining this division — is done by an artificial intelligence, which then decides not to show the latter, we may escape the problem … but can it do so if it were human-like?)

    However, it may well be that we inevitably will observe the known pathological behaviors by convergent evolution even with a less ambitious goal of creating somethings that can do “any job a human currently does”.

    I would not dwell upon that matter, if it weren’t really interesting to me, in particular what you mean by:

    The hope is that we can connect the genetic and molecular perturbations to the circuit dynamics and then connect the circuit dynamics to behavior. In this sense, we can circumvent the really hard problem of how the canonical circuits are connected to each other. This may not lead to a complete understanding of the brain or the ability to treat all disorders but it may give insights into how genes and medication act on cognitive function.

    [my highlighting]

    Is this the idea that convergent evolution lets you bypass the hard problem?

    Like

  7. @Markus Interesting points and I don’t have the answers. The mental illness question is a good one although I agree that the distinctions are quantitative and not qualitative. I believe that machines can do what humans can “on average”. There will always be outliers that are so unlikely that the probability of finding that parameter space is vanishingly low. I’m not really relying on convergent evolution to bypass the hard problem. What I mostly think is that some illnesses are due to low level hardware problems and we can at least try to elucidate those in terms of low level functions that can be measured.

    Like

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