Musings of Navigating The Finite remainder of life from Porchville, with the hope of a glimpse of The Infinite

Sunday, July 3, 2011

The Future of Human Intelligence

I have turned into a bit of a neuroscience junky after having read David Eagleman’s  Incognito.  The other day I was browsing about the web and ran into the description for an article in Scientific American on The Physics of Intelligence.  Cool, after reading about quantum computers I needed to return to a machine that operates on classical physics, our brains.  Unfortunately to read the article will require a subscription.  I frittered away a good hour arguing with myself as to why or why not I wanted to subscribe—ergo proving to myself one of Eagleman’s contentions that our minds are a collection of competing individuals.  The CFO (chief financial officer) said no--too expensive.  Human resources said no—you don’t have time for another subscription.  The manager of past experience said no—you have subscribed to Scientific American before and found it too rigorous.  Most of the members of the board of directors of Sextant Inc. (at least the ones that were not napping through the board meeting) voted no.  The CEO, the real me, the me of the moment, the me that is in “control”, the me that is the President of the United States Of The Mind of Sextant said “to hell with all you cheap bastards I want to read this article and I am going to subscribe.”  Which is pretty much where I had been 1 hour previously. 

With all these deliberations over subscribing to a frigging magazine, you can imagine the mental conflagration that raged in my mind over my decision to retire or not…which by the way we are now entering into month three.  Hmmmm!  It seems like it has been three years since I have worked.  Do I miss it?  Not in the least.  I am however missing my friends and the social contact that I had at work.  But work itself?  Nada!  I love retirement, just wished the damned clocks would slow down.  They have been running in high speed weekend time since I retired and my life seems to be running in fast forward.  Maybe I should be careful about what I wish for! 

Getting back to the Scientific American article, it is titled The Limits of Intelligence by Douglas Fox.  Again you can’t read the article without a subscription but here is a link for purchasing the July 2011 issue:

Ahhh, haaa! All is not lost, here is a 26 minute podcast conversation with Fox discussing the article.  You will get the gist of the article by just listening in--for free:

You know how every once in a while you read that evolution is not a static thing, that human beings are still evolving, and often there are speculations about Homo futuramus having a 13 pound brain and making Einstein look like a dunce.  Well Fox brings up some very interesting physical problems that evolution will have to face.  Oddly enough they are very much the same problems that faced Seymour Cray when he was designing his supercomputer back in the ‘70s. 

The obvious solution is to make the brain bigger, something that women may not appreciate, but beyond trying to fit bigger brains through a birth canal, there is another problem.  A cow is no smarter than a mouse, yet its brain is a hundred times larger.  Well much of that is due to the fact that there is a lot more cow than mouse for the brain to operate.  But there are other factors as well.  The number of pounds of mush in our heads is not what makes us smart.  It is the connections, called synapses, and the speed at which those connections operate that instill intelligence.  So big brains have three problems: 1) they consume a lot of energy, 2) they require longer wiring for connections, and 3) because of the longer connections, they have a slower signal speed.  Compounding that problem, the addition wiring takes up more space adding to the size problem, and the higher energy consumption requires greater blood circulation.

So why not make the neurons smaller and pack more of them and their synapses into a tighter space?  Let’s do to neurons what Intel did for transistors back in the early 70s with the integrated chip!   Excellent idea!  The only trouble is nature already thought of that one.  Primates, and humans being the biggest brained primate in town have neurons that are fundamentally different than the rest of the animal kingdom.  Our neurons are compact and extremely well connected compared to the other animals.  Fox gives an amazing comparison.  If a rodent’s brain was to match the number of neurons and synapses of a human brain using the usual rule of scaling, the rodent superbrain would weigh 45 kilograms (99.2 pounds).  So our notion of packing the brain more densely with neurons is such a good idea, that evolution employed it several million years ago.  You know Mother Nature didn’t fall off the turnip wagon yesterday. 

Well hold on now, an Intel Core i7 has certainly evolved leaps and bounds over the 8088.  Why can’t our brains do the same thing?  Well first of all they already have,  remember 99 pound mouse brain.  But further improvement is not a matter of using evermore refined masking technology like Intel does.  First evolution don’t work like a chip factory, it doesn’t quit making 80486 processors because it just designed the Pentium.  Dig down into our brains deep enough and you find the biological equivalent of the Eniac, the 30 ton Neanderthal that used vacuum tubes in World War II.  

Built on top of that are ever more sophisticated structures for higher functions.  Eagleman said that evolution is a tinkerer not engineer.  So it makes improvements but in an additive mode, piling the new innovations on top of the old junk which somehow has to all work together, not in a replacement mode that the computer industry uses…shit can that old laptop and try this new whizbanger.   The other problem is that our brains and the constituent neurons are not these lovely well engineered circuits precisely lasered onto pristine sheets of pure silicon at nanometer scales with 4 atoms of precision.  No most of us come about by mom and dad doing the horizontal mambo after a heavy evening of drinking.  The neurons are electro-chemical goobags that arose out of the primordial ooze.  Yep get us down to our smallest parts and we are nothing but tiny hunks of organic mush housed in a semi-permeable membranes.  What the design of neurons has going for it is a lot of time and the hard knocks of evolutionary experience. 
Neuron, showing axon and synapse. 

The problem with neurons lies in the axon, the long tail that conducts the message from one neuron to the next.  Make a brain bigger and the axon has to get longer.  Long skinny axons transfer signals slower, so the axon has to increase in thickness.  The rub comes in from the fact that if you double the diameter of the axon, you double its energy consumption, and yet only benefit a 40% increase in transfer rate.  The larger axons add to the bulk of the brain making it even larger.  So bigger brains using conventional neurons get caught up rather quickly in a vicious cycle of diminishing returns of more bulk and higher energy consumption canceling any improvement in quantity and transfer speed.  This is why cows are no smarter than mice, and why one would end up with a 99 pound rodent brain to match human performance.

But the new and improved primate neurons are not without axon problems as well.  Like an Intel Core i7 the design is to keep them small with a lot connections and pack them into the tightest space possible.  You want small computer chips, you have to pack the transistors tightly and the conduction paths and surrounding insulation has to be extremely small. 

Intel Core i7, 45 X 42 mm
781 Million Transistors

The Core i7 uses 32 nanometer technology.  That refers to the average half pitch of a memory cell, and what the hell that means is beyond me other than the fact that it is extremely small.  We are talking 32 billionths of a meter.  To give you idea how small, the range of diameter of a human hair is 17 to 180 micrometers (millionths of a meter), 1000 times larger.  But how about neurons, the human kind, the best that Mother Nature has whipped up in her laboratory of evolution?  Well according to Eagleman, just tossing conversational figures around in Incognito:

"The cells are connected to one another in a network of such staggering complexity that it bankrupts human language and necessitates new strains of mathematics. A typical neuron makes about ten thousand connections to neighboring neurons. Given the billions of neurons, this means there are as many connections in a single cubic centimeter of brain tissue as there are stars in the Milky Way galaxy." *

He gets a bit more specific later:
"In a single cubic millimeter of brain tissue, there are some one hundred million synaptic connections between neurons."**

The BB has a volume of 43.7 cubic millimeters.
A BB sized hunk of brain  has 4.37 billion synapses.

So of course you are wondering, well, just how big is a cubic millimeter?  It is a square box 1 mm X 1 mm X 1 mm.  It is not microscopic, you wouldn’t have any trouble seeing one setting in the bottom of a teaspoon.  To fill that teaspoon you would need 4,929 cubic millimeters.  Perhaps you can better visualize a cubic inch.  Imagine a cube 1 inch X 1 inch X 1 inch.  It would fit nicely in the palm of your hand.  Now get a pair of imaginary tweezers and start filling the imaginary cubic inch with the imaginary cubic millimeters.  Let’s suppose that you can place one cubic millimeter into the cubic inch per second.  This task is going to take you, about four and half hours non stop, because there are16,387 cubic millimeters in a cubic inch.  Still can’t get an idea?  Remember BBs from the BB guns of your youth?  A BB has a diameter of 4.37 millimeters (0.172 inches) and has volume of 43.7 cubic millimeters.  So using Eagleman’s estimate, there are 4.37 billion synaptic connections in a hunk of brain tissue the size of a BB.  As I mentioned earlier Mother Nature didn’t fall off the turnip cart yesterday.  So you want to improve on that, have at it, but good luck! 

Remember that axon?  The wall is a membrane and it has what are known as ion channels.  Remember a neuron is an electro-chemical device, probably more related to a battery than a transistor.  These ion channels open up and let positive charged ions into the axon.  When you get enough ions in the axon, the potential will build up and zap…the neuron fires.  But what happens when the axons get real small?  The same thing that happens when you start getting into progressively smaller nanometer architectures in computer chips, current leakage becomes a problem.  Neurons are goobags remember, and a certain number of ions will leak through. The smaller diameter axons are more susceptible to unwanted firings due to ion channel leakage.  So tiny engenders unwanted signals, called noise, in both computer chips and neurons.

So what happens when axons short out or become excessively noisey.  Well there is a subject that I, and anyone with MS, has first hand knowledge.  Multiple Sclerosis is the removal of the myelin sheath on the axons by one’s immune system.  It is like stripping off the insulation on wires.  The neurons short out.  You send a command to move your arm.  The thousands of nerves that control the muscles in your arm are shorted out.  Your arm does not move or moves with very little strength.  So I can vouch for the fact that you don’t want to mess with axons. 

The last problem which I have alluded to but not covered, is that any improvement in intelligence will come at a high cost in energy consumption.  Right now our brains at rest consume 20% of the oxygen usage and caloric input to our bodies.  At what point would more intelligence tax the rest of the system in a deleterious fashion? 

So the long and short of all this is that many neuroscientists believe that the brain is near the end of nature’s ability to increase intelligence.  Any improvement creates other problems which either cancel the improvement or provide diminishing returns for the energy investment.  We may not be able to get much smarter than we already are. 

Is the approaching end of evolution of intelligence a cause for concern?  Well I don’t know about you, but I am not going to lose any sleep over it.  First of all, we would never live long enough to observe an improvement.  Individuals do not evolve, generations do.  Noticeable improvements would occur over millennia, not life times.  So to me it is sort of like worrying about the sun consuming the Earth when it becomes a red giant…we will all be destroyed!  Yes in 5 billion years. 

So are we stuck with remaining the same old stupid beings for the foreseeable future?  Not in the least.  A normally intelligent person today knows far more than the smartest genius of 600 years ago.  There is not a person alive today who could disavow Issac Newton’s statement “If I have seen a little further it is by standing on the shoulders of Giants."  None of us could survive on the gift of our intelligence alone.  Like the neurons in our brains, our strength as a species comes from our ability to communicate with not only each other but with the generations of the past and to the generations of the future.  Twentieth and twenty-first century mankind will not be remembered for any genetic improvement in the DNA of the species—at least at this juncture.  It will be remembered for the technology that it bequeathed to the following generations.  A student struggling with long division on a yellowed ruled tablet is not going to get any better answer to a quotient than another kid doing it on a calculator or computer.  Yet learning how to effectively use the calculator or computer, may be a far better use of time than scrawling out figures on paper. 
The Layers of Time

The Long Now Foundation has an interesting graphic called the layers of time which is shown to the right.  I have often thought about this without seeing it so conveniently illustrated.  At the bottom of the graphic is nature, and here is where our physical evolution occurs.  The next level up is culture, then governance, infrastructure, commerce, and finally fashion and art.  The gist of this graphic is that the layers at the top of the graphic move at a faster pace through our lives and has a more facile affect on humanity than the lower layers.

The School of Athens, Raphael

We are not genetically much different, if at all, from those who lived during the time of Socrates and Plato.  Let’s do one better than Raphael, imagine magically gathering the cast of characters depicted in his School Of Athens and dropping them into a meeting of the United Nations General Assembly.  These people would be utterly flabbergasted by our technology, yet would have no problems understanding the philosophical, political, and cultural underpinnings of all of our problems once they were introduced to the modern issues and players.  Given a month or so to learn the ropes, I would image that Socrates or Plato could embrace modern life with little difficulty.

United Nations General Assembly in New York

We don’t change that much over time, nor will we in the future, but our technology and what we can do with it is vastly different from the time of Socrates and will be vastly different in the future.  What bridges all of these layers of time are language, written literacy, science and technology, and education, and it is with these tools, not the physical evolution of the brain, through which we will continue to evolve into realms that we can only imagine right now.     

Image Credits.

Cover, Scientific American:  Scientific American -- Digital, July 2011

The Layers of Time:


*Eagleman, David (2011). Incognito: The Secret Lives of the Brain (pp. 1-2). Pantheon. Kindle Edition.

**Eagleman, David (2011). Incognito: The Secret Lives of the Brain (p. 173). Pantheon. Kindle Edition.


  1. What is intelligence? It turns out that the answer depends on whom you ask, and that the answer differs widely across disciplines, time, and places. We discuss the diversity of views about what intelligence is because empirical studies often assume rather than explore the nature of the construct they are investigating .in this case, intelligence.

    Distance Calculator

  2. Excellent point! The article did not attempt to define intelligence which is a slippery slope that can lead to despair. I remember reading something about 10 or 15 years ago about considering intelligence in a much wider context than one's performance on an IQ test, and the resulting branding that one must bear for life by having an ill defined parameter documented on one's permanent record. The article considered the intelligence of passing a football during an NFL game. The article mentioned Joe Montana and I believe that it said that Montana's terse (at the time) demeanor could be attributed to the "football passing computer" borrowing areas of his brain devoted to speech. Eagleman, in Incognito, mentions that when faced with a new task, for instance riding a bike, we have a very difficult time with it until we burn the methods into our unconscious mind (the unseen portion of the iceberg). At that time such tasks become unthinking algorithms that run automatically. Riding a bike is not going to show up on a IQ test, yet it certainly is a complex form of "intelligence".

    If one wishes to avoid the "I" word, then I suppose you could say that the gist of the article is that an improvement in those qualities and abilities that are exhibited by having many fast synaptic connections are near the end of the evolutionary rope due to the physical trade offs providing diminishing returns in performance.

    Googling around I ran into some interesting information on "multiple intelligence". Alas I remain a dunce in all categories! I have a theory on that...the thing that I could really excel at has not been invented yet.

    Again excellent point. Thank you for commenting.

  3. What I read about Joe Montana was a subtopic of The Mind's Sky by Timothy Ferris, "Joe Montana's Premotor Cortex" (on page 99 which naturally is not part of the viewable content).

    Ferris also wrote a newspaper article regarding the matter: