Here's an article about HP's plan for next generation computing built around memristors (dreamed up by Amy Chua's dad 40 years ago). The problem with the current architecture:
Today, computers constantly shuttle data back and forth among faster and slower memories. The systems keep frequently used data close to the processor and then move it to slower and more permanent storage when it is no longer needed for the ongoing calculations.
In this approach, the microprocessor is in the center of the computing universe, but in terms of energy costs, moving the information, first to be computed upon and then stored, dwarfs the energy used in the actual computing operation. ...
One reason is computing’s enormous energy appetite. A 10-petaflop supercomputer — scheduled to be built by I.B.M. next year — will consume 15 megawatts of power, roughly the electricity consumed by a city of 15,000 homes.
There are perhaps analogies here to the evolution of human intelligence. For millions of years, our predecessors' brains got larger, peaking with the Neanderthals. That suggests that bigger brains made us fitter.
But that's a brute force solution.The usual argument is that big brains require too much food. (They also make us fall over more.) Now, it could be that as human population increased after the invention of agriculture, there were enough mutations to create more intelligence per cubic centimeter, just as there have been with computer chips.
That's certainly true, but I suspect the concomitant problem of not just getting enough energy in, but of getting enough energy out of the skull, of heat dispersal, also became a problem with this trend toward bigger brains. With a roughly spherical shape, as volume goes up, so does the volume to surface ratio.
Unfortunately, I've never seen anybody who actually knows what they are talking about consider this question of brains shedding heat.
It could be that there is an ideal latitude at which the cost of keeping the brain warm is balanced by the cost of keeping the brain cool at lowest overall cost. In 1911, Yale Professor of Geography Ellsworth Huntington conducted a study of climate's effect on human achievement. He concluded that the ideal climate was roughly that of New Haven, Connecticut. In a recent article, Malcolm Gladwell had great fun with that: here we are, 100 years later, and we can see what a biased moron Huntington was! Proving how much things have changed in 100 years, Malcolm's article appeared in that glossy, ad-packed magazine, The Lagoser.
Posts
28 comments:
If they take full advantage of the memristor, they could restore computing to the analog realm. If that changeover is done properly and fully, electronic brains could truly begin to compete with squishy brains.
Sounds like HP isn't really switching modes but just using memristors as a way of packing more digital into the same space.
Wow, IBM can get over 650 Million floating point operations out of a single Joule of energy! For a single penny of electricity at industrial prices (approx. 6 cents per kilowatt-hour), we're talking about 400 Trillion operations - 60 thousand for every human. It would take a 1-Megaflop computer (which would have stunned folks as almost magical only 40 years ago) an entire decade to do that.
For a penny.
And this is considered a big problem the solution for which requires exotic cutting-edge memristor technology.
Since big brains use so much energy, a big head is a sign of conspicuous consumption, therefor proof of fitness and sexual selection. Like a peacock's tail.
Since brains are useful for out-bullshucking members of your clan, and by extension for planning ahead to get food and shelter, big brains are proof of fitness and sexual selection.
I think this overexplanatory Just-So side to sexual selection 'it's useless, so it's sexual selection' and 'it's useful, so it's sexual selection' is why Gregory Cochran hates sexual selection. It's inelegant. But I think it's true.
"With a roughly spherical shape, as volume goes up, so does the volume to surface ratio."
Hm, this could explain why some of the smartest cat breeds have flatter appearing skulls and huge ears: the Devonshire Rex, the Bengal, the Russian Blue, the Siamese, etc.
While most of the waste heat in the brain is carried to the thorax by circulating blood, our nasal passages are also a very effective heat exchanger. (No, you don't lose 75% of your body heat from your scalp: http://www.wintercampers.com/2011/02/19/the-truth-about-heat-loss-through-your-head) Perhaps the big noses and cavernous sinuses of Europeans and, ahem, Jews allow their brains to operate cooler and more efficiently. So while Asians increased operational speed and complexity with incrementally bigger organ size, Europeans grew a bigger proboscis. Also, Westerners have larger foreheads and thinner hair, which might give them a brain cooling advantage.
What's next in homo sapiens evolution: a big fin on our head or a lolling, grotesquely large tongue for heat exchange?
When it comes to re-creating human intelligence computeristically, maybe scientists need to think in terms of consciousness or awareness first. Even the most powerful computers are not conscious. A goldfish can't do math or understand language and has very limited memory, but it has awareness, crude though it may be. A computer can defeat man in chess and do all sorts of awesome calculations, but it has no awareness.
To be intelligent means to be able to consciously use one's abilities. Computers are logical or algorhythmical--whatever that means--but they are not intelligent as long as they have no consciousness. Otoh, a newborn baby is conscious even though it has virtually no intelligence; it can't even add 2 + 2.
How may scientists create artificial consciousness or sentient-ness? Maybe there is a 'spontaneous' law of life, something like the force of gravity. Gravity is something we don't fully understand, but it's there in every bit of matter. There is a natural 'desire' of matter to come together. And when enough matter comes together, it forms a solid mass, and when even more matter comes together, it heats up and creates something like the sun. Each hydrogen particle by itself aint much but if gravity brings together super-tons of that stuff, they 'spontaneously' heat up and turn into a sun. They light up. So, instead of just trying to figure out cognition, we need to maybe think about ignition.
Maybe like the law of gravity, there is a natural 'desire' on the part of life particles to connect with one another. Maybe we won't fully understand this dynamic. Understand it or not, suppose such a force exists. Each neuron, just like each hydrogen particle, doesn't add up to much, but maybe if you put a whole bunch of neurons together, there is a 'spontaneous' process whereby the neurons began to connect and communicate with another. And if there are enough neurons, it ignites into awareness--the primitive kind of a fish--and then if there are more neurons added to it, then it expands into consiousness.
So, maybe, in order to create artificial consciousness, what scientists need to do is to create a cybron--a cyber neuron--and then electronically connect a whole bunch of cybrons together, and then just sit back and watch as the 'natural' process takes hold. Maybe 'spontaneously' the cybrons will start interacting with one another and create a kind of awareness. And the more cybrons we add to the system, the more consciousness it will be. If this happens, we can then add 'senses' to what is like an artificial Helen Keller--like lens for eyes, and microphone for ears and feed it lots of data. I know this aint science but maybe it has value at least as sci-fi.
Steve, I don't think that the analogy that you are after (ie of a computer chip shedding heat through a 'heat sink'), is the correct one with regards to human physiology.
Yes, cells 'burn' energy i their metabolism, and that energy creates heat, yes there is an optimal temperature for cell biology in warm blooded animals to operate - too hot = death, too cold = death, but the 'problem' really cannot be compared to the familair mechanical problem of a heat source emitting damaging amounts of haet, and a raidiator system being used to dissipate that heat.
For starters, the body temperature of humans must be kept by the physiology at a constant 38.9 C, or whatever it is, within very tight bounds.Ambient temperature is rarely higher and the body has some very good cooling mechanisms (eg vaso-dilation, sweating), to shed heat if warmer.
To think of the skull as some sort of insulater, stopping the brain from 'shedding heat' is also wrong.As you know, all internal parts of the body are permaeted and bathed in fluid, blood or otherwise.This is circulated at a phenomenal rate, distributing the heat.
Think of how large an elephant's brain is.They are warm blooded and live in hot countries, their brains operate fine, so is there a size/heat limit?
"Unfortunately, I've never seen anybody who actually knows what they are talking about consider this question of brains shedding heat."
Common baldness?
Lots of apes bald on their heads. One kind of little ape, the macaque, almost universally balds, both male and female.
Scientists sure seem to go bald more than other people. Maybe all that headache-inducing deep thought causes their brains to stay hotter longer, and their dna gins up a cooling solution for them over time? Mechanically it would make a little sense?
Bigger heads? Larger nostrils? Bigger ears? Larger foreheads?
Throughout the animal kingdom, instruments of male territorial aggression have been selected for in race conditions that produce some of the more puzzling phenomena. Envision the brain as a weapon used in sexual and territorial competition between males. Instead of using horns, antlers, or fangs to best our opponents, we use tools and strategies - products of the mind.
"Unfortunately, I've never seen anybody who actually knows what they are talking about consider this question of brains shedding heat."
Dean Falk has written on this issue. In a nutshell, her hypothesis is that you can only have a brain as large as you have the capacity to keep cool. Follow this link and scroll down to "The Radiator Hypothesis."
I became somewhat entranced by the radiator hypothesis some years ago and began to wonder if that was the reason humans had lost their fur - i.e. by turning our entire bodies into heatsinks we enabled the growth of a larger brain. Of course that raises the question of how we can avoid hypothermia without fur. So I began to wonder if the taming of fire had released our ancient ancestors from the hypothermic constraint thereby enabling them to shed their fur and in turn grow larger brains.
I wanted to find out if the advent of fire coincided with an increase in brain volume. However, all I found out was that the date of the domestication of fire is extremely uncertain. So I really couldn't come to any conclusion.
"Since big brains use so much energy, a big head is a sign of conspicuous consumption, therefor proof of fitness and sexual selection. Like a peacock's tail."
Reading this and the comment on common baldness has got me wondering...
...how come this macrocephalic chrome domed stud hasn't been GETTING MORE?
Gilbert Pinfold.
"Proving how much things have changed in 100 years, Malcolm's article appeared in that glossy, ad-packed magazine, The Lagoser."
That was funny!
Actually, head and body temperature optimum considerations have not really had much bearing on evolution for a long time, since people have been adjusting to the weather with clothes as needed for thousands of years...
You can keep putting clothes on as it gets colder, but you can't keep taking clothes off as it gets warmer.
"So, maybe, in order to create artificial consciousness, what scientists need to do is to create a cybron--a cyber neuron--and then electronically connect a whole bunch of cybrons together, and then just sit back and watch as the 'natural' process takes hold. Maybe 'spontaneously' the cybrons will start interacting with one another and create a kind of awareness. And the more cybrons we add to the system, the more consciousness it will be."
DON'T YOU DARE!
Good God, man, poor Arnold Schwarzeneggar devoted almost 2 decades to defeating Skynet, and you want to bring it all on, AGAIN?
"Steve, I don't think that the analogy that you are after (ie of a computer chip shedding heat through a 'heat sink'), is the correct one with regards to human physiology.'
then how about liquid cooling?
You can keep putting clothes on as it gets colder, but you can't keep taking clothes off as it gets warmer.
As it gets really warmer, the solution is to put more clothes on. Turbans and heavy insulating robes worn by people in deserts are for this precise purpose.
Perhaps the big noses and cavernous sinuses of Europeans and, ahem, Jews allow their brains to operate cooler and more efficiently.
What about Arabs?
"You can keep putting clothes on as it gets colder, but you can't keep taking clothes off as it gets warmer."
I just made that exact point here. Somewhat-above-average minds think alike.
In Terry Pratchett's "Discworld" the normally thick-as-a-brick trolls (made out of rock) become smart at very low temperatures.
"Perhaps the big noses and cavernous sinuses of Europeans and, ahem, Jews allow their brains to operate cooler and more efficiently.
What about Arabs?"
Controlled experiment: have plastic surgeons in Beverly Hills administer an IQ test to all the Arab women who come in for nose jobs -- then test them five years later when they come in for liposuction.
As it gets really warmer, the solution is to put more clothes on. Turbans and heavy insulating robes worn by people in deserts are for this precise purpose.
Only if its dry. In humid climes, putting on more clothes makes you hotter.
"You can keep putting clothes on as it gets colder, but you can't keep taking clothes off as it gets warmer."
But you can invent and use air conditioning, and maybe the mind developed a kind of bio-air-conditioning.
The brain cooling idea doesn't make much sense. From this link, humans have 1.4kg brains, elephants have 4.8kg brains, and sperm whales have 7.8 kg brains. Elephants tend to live in rather hot places, so there's pretty substantial evidence right there that cooling isn't the limit on our brain size. (It's probably one limit on total body size of land animals, but I think bone strength imposes a constraint before cooling ability.)
From another angle, we know what it looks like when your brain and body overheat--you get heatstroke. That happens to healthy adults pretty-much only when they're physically very active in hot temperatures and don't drink enough, or are stuck in some super hot confined place--it's not something that happens to you from thinking too hard on a hot day.
I think the practical constraint on human brain size is fitting that head through the birth canal. You don't see substantial numbers of healthy people with big heads dropping dead from heatstroke, but you do see substantial number of healthy women having C-sections, and while some are probably unnecessary, a lot of those women would have died in childbirth without surgery. That looks like bumping up against a constraint--bigger brains pay off, but not if mom can't get the baby out.
I'm afraid that the limiting factor in human intelligence was not heat dispersion, it was the drawbacks to intelligence itself.
A duck is just intelligent enough to be good at mating and finding water bugs, but not enough to question why he should fly south for the winter.
There appear to be a number of intelligence governors that keep people from questioning "irrational" behaviors that enhance their reproductive odds.
In fact, it is no accident that racial groups generally display an inverse relationship between IQ and birthrate.
"The brain cooling idea doesn't make much sense. From this link, humans have 1.4kg brains, elephants have 4.8kg brains, and sperm whales have 7.8 kg brains. Elephants tend to live in rather hot places, so there's pretty substantial evidence right there that cooling isn't the limit on our brain size."
Ever notice the brain heats up when it's thinking a lot. Whales and elephants have big brains but don't think a lot, so maybe they don't heat up as much. Also, whales live in the cool ocean. And elephants have ears as fans and trunks to spew water on the head with.
The idea that you burn alot of energy and generate alot of heat in thinking is not correct. When I was a biology undergrad, I took a senior level course in biochemistry where one of the questions on a test involved calculating the amount of energy spent by the brain in taking a test of roughly the same length of time as what we were taking; the prof provided the necessary facts to us to allow us to figure it out (the missing data was information on glycolysis that we were expected to know from the course). Turns out, it is much lower than what you might think, even working from the assumption that you are going full bore at it.
That being said, the brain is still considered one of the top dogs by the body- nutrients will preferentially get shunted there by your body in the event that you are starving.
Also, the brain is vulnerable to environmental heat and cold- the body has a number of mechanisms for heating and cooling such as sweating, cappilary dialation, etc to ensure that the brain doesn't get too hot or cold.
I imagine that Northern Germany has a climate very similar to New Haven.
>this overexplanatory Just-So side to sexual selection<
Sexual selection is where it's at, so a lot of energetic variations and experimentation (er, so to speak) occurs in the heated and puzzling realm of what people find attractive. Big butts - slender butts - longer hair - shorter hair - pale - dark...one almost hears Nature barking in a frenzy of lust.
Sometimes the heat of the computer screen heats up my forehead, and I have to go outside (in winter) or stand under a fan (in summer) to restore my normal temperature. This may be a factor. Anyone experience brain heating in the absence of nearby electronics? (I sit pretty close to the screen.)
Post a Comment