Debunking the Brain/Computer Connection

Many thanks to CM who sent a link to me for the online version of the WSJ article I referred to right before I left for the LYC Teen Camp.

Like Mr. Lee Gomes, the writer of the article, I am frequently annoyed by the common comparisons made between the human brain and a computer, as if there really were any comparison. Mr. Gomes dubs it the Synapse Equivalency Fallacy. I am equally nauseated by the related claim that we have dispelled the mystery of the human brain, that we have it essentially nailed down… After all, they say, we now know that the “mind” is but an illusion and that what we think of as consciousness and will is merely a mixing of chemical and electrical impulses.

It’s nauseating because it’s just not true, and I’ve never heard or read a study or article that could truly be extrapolated to such an extreme conclusion. After decades of mapping and analysis, the brain and its relationship to consciousness and will is still an incredible mystery. I’ve commented on this topic before (here, here, and here), and the nature of the brain and mind are favorite interests of mine.

[FWIW: Personally, I believe that the human mind arises from the combination of the human brain and the human spirit (e.g., Job 32:8 ) as two essential ingredients. That the mind is not purely a product of the brain alone is what I believe helps to explain so-called “downward causation” and phenomena such as discussed here.]

But the focus of the WSJ article focuses narrowly on annoyance number one: the idea that if we simply pack in enough computing power we will have the equivalent of a human brain. As the article says, the comparison between brains and computers is simply apples to oranges: there really is no good correlation beyond the simplest of analogies. In fact, even the simplest of animal brains still confounds us. I found this particular paragraph in Mr. Gomes’ article just delightful (emphasis mine):

Dr. [Christof] Koch [a professor of biology and engineering at CalTech] tells a cautionary tale involving Caenorhabditis elegans, a tiny, soil-dwelling worm famous in science for being one of the best-understood animals on the planet. The brain and nervous system of C. elegans contain only a few hundred neurons, and by now all of them, along with all of their interconnections with one another, have been charted out in great detail. “We have a complete wiring diagram,” he said, “But we still have no idea of how it works; no overall view of how the animal functions.”

Dr. Koch then says this is “even truer” with the human mind and that there aren’t “any good theories” on how the physical brain is able to do all the we know the brain does, “including keeping us conscious” — causing me to think that the next time I hear someone speak as if we really understand something about the human brain or mind, I may declare as a battle cry: “Remember C. elegans!”

The final few paragraphs are a worthwhile summary of the article’s thrust:

But isn’t Moore’s Law on technology’s consistent doubling of capacity going to give us, in not very much time, computers with so many transistors they’ll have the same processing power as the brain?

Yes, said Dr. Koch, but absent a better understanding of the brain, that hardware will be of little use.

“With bigger computers, all we are going to get is more junk. It would be like someone in 1900 saying, ‘Give us more slide rules and we will understand the universe.’ But they would have been saying that without knowing of Planck or Einstein or Heisenberg. And they would have had no idea of what they were talking about.”

(Any day slide rules are mentioned in the newspaper is a good day…)

The entire article is worth a read, and includes a video of the article’s author, WSJ columnist Lee Gomes. Check it out: “Linking Brains, Computers”.

4 thoughts on “Debunking the Brain/Computer Connection

  1. Even if computer software could somehow replicate the information processing capacity of the human mind and its intelligence (and no one is anywhere near that), such a computer would still not be conscious. Even if you could make a computer-controlled robot that could mimic a human being so closely that an observer could not tell the difference without opening it up, it would not be conscious. It would not feel things as we do. It might simulate crying or reactions to pain, but it would not be cruel to destroy it.

    Scientists are so far from addressing the issue of consciousness that they write as if they don’t know what it is. Which is understandable, since there is nothing science can say about it except “we don’t know where it comes from or how it works.”

    How ironic, that science which prides itself on knowledge and its methods of acquiring knowledge, can say absolutely nothing about the most obvious, self-evident fact for every human being on the face of the earth, “I exist and am conscious.”

  2. rakkav

    It is interesting all the same that artificial intelligence is already at the point where in specialized areas, it can far outperform the human brain (e.g., in chess and in checkers — the latter problem now “solved” so that the best any opponent, human or machine, can do is play to a draw, as I noted in a recent blog of my own).

    Of course, the same argument may be persuasively made about machines that can lift loads many times those within human capacity to lift. But it doesn’t prove there what some people think artificial intelligence proves about the human brain. Any machines that man makes, including artificially intelligent ones, are but extensions of his own brain and body and are thus inferior qualitatively to them (even if superior quantitatively — that is, after all, their point).

    It’s amazing how few people seem to think this matter through correctly. Maybe *theirs* is artificial intelligence (or is it natural stupidity?), or maybe they’re now part of the Pod People. Yes, that must be it: Satan’s Pod People. He has a deception for everybody, including you and me if we let him hand us one.

    ברכות בישוע המשיח
    יוחנן רכב

  3. Howdy, Mr. Wheeler, and thanks for your insight. I think you make a good point, but I don’t know that we would agree that we have created anything even quantitatively superior to the human mind without some qualification.

    For example, the “championship” nature of the chess and checkers computers boils down ultimately to their ability to calculate & evaluate an amazing quantity of possibilities at amazing speed (like Superman on an abacus). But their focus is so narrowed, that I don’t know if they could do any other calculations whatsoever without reprogramming by a human being. Can they add up the years to determine the age Frederic will be when he is finally able to marry Mabel in the Pirates of Penzance? My kids can figure that one out in a few moments (for some of them, more moments than others).

    I know it’s picky, and it really doesn’t invalidate what you are saying. My differences surely fall under the “qualitative” categorization by most measurements. But I am still hesitant. I don’t like referring to the “creator is of necessity greater than the created” argument too quickly, because in too many circumstances it seems too easy a concession and an argument that suffers from too much apparent weakness (though, I would say not actual weakness).

    The fact is that artificial intelligence efforts have not even come close to replicating real human intelligence, nor — seen in proper perspective — are they even knocking on the door. Nor are they in the neighborhood. Nor the city, state, country, or — apparently — planet. (Extend as you wish: solar system, galactic arm, etc. Fuzzy analogies are fun for their lack of bounds!) And should an AI researcher parade a computerized checkers champion as a sign that gleeful barbarians are at the gate, I would suspect that he needs a perspective check.

    God didn’t design our minds to maximize chess victories, and it should not surprise us that we can create machines that can win more matches than we can. Interestingly, if we really could create a computer that successfully and fully replicated the human brain, not only would it be one of the single greatest technological achievements of all time, but we would have created a machine that is poorer at winning chess matches than today’s computerized chess “champions”. Understanding this, I think, highlights the incredibly vague point I have been trying to make here. 🙂

  4. When I was a little boy, my accountant grandfather taught me how to use a calculating machine. You punched in the numbers, then pulled the hand lever. The little roll of paper gave you the answer.

    Today, computer scientists try to mimic the synapsis in the human brain with “neural network” programming. Problem is, the program can only follow the pre-arranged structures inputed into it.

    When a computer develops independent thought, introduces subjects for which no programming exists, only then will I take artificial intelligence seriously.

    A computer is nothing more than a fancy calculating machine.

What are you thinking?

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