Eight arms and nine brains

 kw: book reviews, nonfiction, octopuses, intelligence, philosophy

After reading (and reviewing) Metazoa by Peter Godfrey-Smith, I just had to get his earlier book Other Minds: The Octopus, the Sea, and the Deep Origins of Consciousness.

The author has spent much time at a place near Australia called Octopolis, where as many as a dozen or more octopuses, which are usually solitary, live in rather close association. Near the end of the book he muses about how Octopolis came to be, and mentions in a glancing way a few other reports of these animals living near one another. He has also spent much time "visiting" giant cuttlefish (up to a meter in size) that frequent other areas, also near Australia. These experiences inform his thoughts about "What is it like to be an octopus" and about consciousness in general.

Theories of why we have large brains and consciousness tend to revolve around our sociability. At least among apes, brain size correlates with social environment. So why do octopuses have such large brains? Currently, only at Octopolis are they seen to associate socially, although much of that is fighting. I suppose society has to start somewhere.

The book opens with a quote from William James, which I reproduce here in full:

The demand for continuity has, over large tracts of science, proved itself to possess true prophetic power. We ought therefore ourselves sincerely to try every possible mode of conceiving the dawn of consciousness so that it may not appear equivalent to the irruption into the universe of a new nature, not-existent until then. —William James, The Principles of Psychology, 1890

This exactly accords with my understanding. René Descartes thought that only humans were conscious, leaving him free to unflinchingly torture a dog, claiming the animal's screams were robotic utterances. I titled my review of Metazoa "Is 1% of a mind still a mind?" Let's see if we can figure what proportion of a mind as we know it might reside in the common octopus, Octopus vulgaris.

The nervous system of the common octopus includes half a billion neurons. However, 60% of these reside in eight sub-brains, one near the base of each arm, and the other 40% are in a torus (a donut-shaped mass) that surrounds the esophagus. Thus, the main brain has about 200 million neurons, while each arm's sub-brain has about 37.5 million. We should probably consider the main torus as equivalent to the cortex in vertebrate brains, and the 8 sub-brains, taken together, as equivalent to the cerebellum, which runs the body and, so far as we know, does not contribute to consciousness. The size of the main torus is equivalent to the entire brain of a Norway rat. But if we consider the octopus torus-brain equivalent to the vertebrate cortex, it more closely matches that of an average house cat (250 million) or a less-familiar animal, the rock hyrax (200 million).

Octopuses are mollusks. Of mollusks, octopuses, squids, cuttlefish, and scarcer animals such as the nautilus and sepiolid ("Dumbo octopus") make up the order of cephalopods, which means "head-foot". Among mollusks, all the brains are found among the cephalopods. The largest snails and clams (members of two other orders of mollusks) have brains containing ten to twenty thousand neurons, not millions.

What is the size of the human cortex? We've all heard that we have "about 100 billion nerve cells". The actual number is close to 85 billion, but only 16 billion make up the cortex, where we think consciousness is created, the feeling of what it is to "be a person." Most of the rest are in the cerebellum, which runs the mechanics of our body. If there is a linear relationship at work here, 16÷0.5 = 32: our cortex is 32 times as large as that of an octopus. Does that mean we are 32 times as conscious? Maybe so! Or perhaps self-awareness (which may or may not be the same thing) is a less-than-linear function of brain size, such as a square root. Then we could say we are about 5.5 times as self-aware as an octopus. 

Let's speculate for a moment what it might "be like" to be a honeybee, which has a million neurons. Estimating that half are involved in its sense of self, the little bee may be 1/32,000th as conscious as we are, or perhaps as much as 1/180th. Whatever they think about, it's sure to be a lot slower, and less profound. Beyond that I dare not tread!

Regardless, it is pretty clear from the stories and analyses by Dr. Godfrey-Smith that octopuses are quite self-aware. He tells one story that touched me, of a SCUBA diver who saw an octopus in its den and reached his hand toward it. The octopus reached back and grasped his hand, then came out of the den and began "walking" (they do this on two or four of their arms, with the others coiled or held upward), leading the man on a "tour" of the area that lasted ten minutes. That incident may be the closest a person and an octopus have come to communication on any meaningful level. Far too frequently, the animals are brought into a lab and "tested" with mazes or food choices and so forth. Frequently, they express their frustration by squirting water on the experimenters. Perhaps they prefer not to be treated like the village idiot.

A side thought just occurred to me. The brains of birds and mammals are differently arranged, and the neurons of birds seem to be more efficiently packed and connected. Thus an African gray parrot, with a walnut-sized brain, nonetheless has about 2.5 billion neurons, and can learn to speak and hold conversations with humans. Rather simple conversations, true, but none of us has learned even a single word of "parrot".

As we ponder now we might communicate with residents of a different planet around some other star, we would do well to consider how we can better communicate with parrots, octopuses, and other critters among which we live. 

Do octopuses have a visual language? Their skin is akin to a video screen, and sometimes seems to reflect their thinking. Perhaps much of the time their skin patterns are akin to a visual EEG! I wonder what is on this one's mind.

I really like this author's writing. I've riffed on what he wrote more than reported about it. The best authors stimulate thought!

Intelligence or intellect?

kw: book reviews, nonfiction, philosophy, intelligence, artificial intelligence, history

Short Definitions:

Intelligence – the ability to acquire and apply knowledge and skills

Intellect – the faculty of reasoning and understanding objectively

The first tension that presents itself in Morphing Intelligence: From IQ Measurement to Artificial Brains, by Catherine Malabou, is that between the use of "intelligence" where most philosophers would have preferred "intellect", in discussions of reasoning ability and its measurement, roughly a century ago. I suspect that the definitions she would use for these terms differ from those shown above. Personally, I care little which term is used.

It took me a good while to get used to the language of the book. It was translated from French by Carolyn Shread. I was not always sure if this or that infelicity of English usage was because of the original French idiom, or from word choices made by the translator, choices I might have made differently. I suspect the former; philosophers, particularly European philosophers, think in ways quite foreign to us hoi polloi. Such is the legacy of a century of Linguistic Analysis (both Ideal and Colloquial) and their sister, Logical Positivism: too much musing on "the meaning of meaning". Once I had my mind in gear, I found the book quite engaging.

The core of the book, explaining the title, surrounds three "metamorphoses" of "intelligence." I would perhaps have used either "manifestation" or even "aspect" rather than "metamorphosis". The three are, in brief:

1. The view of Intelligence as a genetic endowment of learning and reasoning ability, that can be measured. It is what IQ tests purported to measure, and in the literature is called both IQ and g, g being the "general factor" as opposed to a seven (later eight) factor model promoted by Howard Gardner. "Mechanical brains" based on digital computers were the early attempts to simulate mental activity in mechanisms.
2. The more nuanced understanding of epigenetic effects, both those which affect DNA expression and those which affect the "wiring" of neurons and synapses. Recent developments of "artificial neurons" such as the "neuro-synaptic processor" or TrueNorth chip (see Cassidy et al, 2016) make significant advances over traditional digital processing.
3. Future developments are expected to result from "removal of the rigid frontiers between nature and artifice." In other words, the "power of automatism" is expected to yield a constructed system that is a brain in every meaningful sense of the term.

The TrueNorth chip, with upwards of 5 billion transistors, simulating the action of a million neurons connected by 250 million synapses, is intended to functionally simulate 250 cortical columns, of which the neocortex of a human brain has about four million. This is based on measurements made since 2009 that the neocortex contains 16 billion neurons, while the cerebellar cortex (which runs the automatic systems of the body) contains about 80 billion neurons. Clearly, the cerebellar neurons are much more locally connected; much of the greater mass and size of the neocortex results from the great number of longer-distance connections via larger-diameter axons.

Assuming that the constructed neurons in the TrueNorth chip can indeed replicate the full complexity of biological neurons, the construction of a full brain simulation would require 16,000 chips for its neocortex, quite a nest of wiring for the "white matter" that ties the neocortex together, and 80,000 chips for its cerebellum…connecting the cortex to what kind of body, I can't imagine just now. So let's look only at the 16,000 neocortical chips. According to the 2016 article, each chip consumes a mere 65 mw. Times 16,000, that comes to just over a kilowatt (1,040 watts). Not bad. That compares well with a computer system reported upon in a short article in Scientific American a few years ago, which was thought to have capabilities similar to a human brain, and required a 9,000,000 watt power plant.

I suppose we can at least estimate that something similar to Moore's Law applies to these systems, with a doubling of efficiency every two years. Your brain and mine each consume about 15 watts of chemical energy (transformed partly to electrical signals). The neocortex requires a third of this, about 5 watts. 1,040/5 = 208, which comes to 7.7 powers of two (doublings). Perhaps the contention of Ms Malabou is correct, and automatism will prevail. Will it do so in the span of 15.4 years, beginning in 2016? That would be some time in the middle of 2031. Maybe I'll find out, because I'll be 84 that year.

All this I push aside in favor of another thought, one not found in Morphing Intelligence: Of what use is a perfect simulation of the human brain? I am reminded of a story by Isaac Asimov, probably found in his collection The Rest of the Robots. Using science as it was known in about 1960, he describes the endeavors of researchers at U.S. Robotics to make a robot "more and more perfect", as directed by the company president. One day the researchers bring him a robot that he cannot distinguish from a "natural" man. Shortly after this, an alien spaceship lands, and in due time, a delegation visits U.S. Robotics. They are shown the new "perfect human" robot. The company president gushes about the huge amount of research and cost required to develop it. An alien responds, "So, what's the point?"

There is indeed a point. Once we understand brain activity and function sufficiently well that we can simulate it perfectly, we have the basis for producing a true AI that is equal to our NI (Natural Intelligence) in power, but different, in ways we can determine. We can produce variations on a theme, perhaps developing new ways of thinking, a mechanism (or a bunch of them) from which we can learn new ways of perceiving and learning and interacting with the universe. Now, that is interesting!

Making chumps of us all

kw: book reviews, nonfiction, psychology, intelligence

How many people do you maintain relationships with? I just checked a few places. I have:

• 91 Facebook "friends"
• 109 LinkedIn "contacts"
• 340 Named "face albums" in Picasa

 I don't use Twitter and I am just getting started with GooglePlus. My son has 983 Facebook "friends" and his girlfriend has 173. If you know about the Dunbar Number, which is 150, and understand its significance, you'd say I may be overdoing it, and my son is definitely overconfident in his social abilities.

Robin Dunbar studied social networks in a range of primates, including humans. Our brain's neocortex is primarily responsible for keeping our social network in order: where we are on the status ladder, whom we owe favors to, who owes us, and so forth. The human neocortex is about five times the volume of a chimp's. Chimpanzee social networks tend to top out at 30, the largest size of a stable troop. Apes and monkeys with smaller brains have smaller social networks.

Do I have an exceptional memory, to be able to put names to 340 faces from my photo archive? Not really. The Dunbar Number doesn't express the size of my lifetime "Rolodex", but the largest number of people the average human can afford to spend time maintaining a relationship with. I don't read every post that shows up in my Facebook news feed. There are about 40 FB "friends" that I actively keep track of. When I am not at my computer, my social network breaks down this way: 

• 45 church members, all of whom I contact frequently
• 16 people at work, in two work groups
• 15 close relatives and in-laws
• 10 next-door, back-door, or across-the-street neighbors

That is an active social network of 86 persons. I suspect that, among my son's 983 "friends", there is a much smaller number with which he actively keeps up. But I know he is more actively social than I am, so perhaps he just represents the next step in human mental evolution!

All this rumination was triggered when I learned of the Dunbar Number in Chapter 26 of David McRaney's new book, You are Not so Smart. The cover blurbs declare, "Why you have too many friends on FaceBook, why your memory is mostly fiction, and 46 other ways you are deluding yourself". True to the promise, the book consists of 48 chapters about the various mental fallacies we all fall prey to. The Dunbar Number is not so much a fallacy as a way to check whether we've perhaps been too greedy in collecting FB friends or Twitter feeds to follow. (By the way, feel free to "follow" this blog. I don't Tweet).

And speaking of blogs, the book is a published embodiment of McRaney's blog. Curiously, when I looked at his list of "all posts", there were 29 items, so he has done some research outside the confines of what he has posted. It also means, you won't actually find everything on the blog, you gotta get the book! Clever dude.

A number of the items he discusses are things we tend to mistake when we don't take time to think. For example, Chapter 25 is titled "The Affect Heuristic", and an example runs like this: Suppose you are offered the chance to get cash for picking red beans out of one of two bowls. Bowl A is huge, with hundreds of beans, many of which are red, while bowl B is much smaller, containing, you are told, a total of 50 beans. The odds are also given: 7% red in the big bowl, and 10% red in the small bowl. You will be given a dollar for each red bean you pick. Most people instinctively select the larger bowl, which the author declares is a fallacy because the odds are against you. But there are two questions I would ask before agreeing to take the challenge. Firstly, how many times can I pick?, and secondly, can I look while I pick? The resulting analysis goes this way:

• One or a few picks (ten or less), and you can't look. Choose the smaller bowl.
• As many picks as you like (or at least 25), whether you can look or not. Choose the larger bowl.

Look again at the smaller bowl, B. 10% of 50 is 5. There are 5 red beans. If you pick a red bean the first time, the odds for the next pick are 4/49 = 8.1%. And the most you can ever earn with bowl B is $5.  I suspect you aren't given the option to look while you pick beans, but if you can pick many times, by the time you have picked out 70 beans or so, you probably have 5 already. Go ahead, pick a hundred times if they will let you! For many of the fallacies and other mental mishaps, it pays to hesitate and ask a telling question.

For others, it seems we are hard-wired a certain way, and quite unlikely to beat the wiring. For instance, there is the Anchoring Effect (Chapter 39). You see something you'd really like, but the price is too high. You wait for a sale. The sale day comes, and you go to have a look. It is marked down by half, but you were hoping for an even lower price. When you begin to leave, an alert clerk offers you another $5 off, and you take it. Did you get a good deal? If it was a Louis Vuitton purse or a leather bomber jacket, initially priced at $800, then marked down to $400, was $395 a good deal? Only if you are accustomed to that lifestyle. Even if you know the price of production is $75, the status may be worth it to you. That's what the premium brand retailers count on. The $800 was an "anchor" to set your expectation, so that $395 seems like a bargain. You can get a great purse that will last you for years and looks good at an outlet store for $60 or less. Ditto a bomber jacket. It just won't have the premium label (unless you buy a counterfeit label online and sew it on yourself). You can take advantage of anchoring, whenever you initiate a negotiation. Start with an unreasonably extreme position, and put the other person in the position of bargaining with you.

The book is full of revealing essays, and nearly all spend a bit of time discussing psychological research that underlines the point. Why do we fall prey to so many fallacies? We are more complex than we think. We have a strong emotional life, which evolved long before the "intellectual" part of the brain. For millions of years, quick recognition, snap judgments, and rapid heuristics saved the lives of our ancestors so they could have descendants. If you had to calculate the weight of that tiger (let's see, nine feet long and two feet wide, so he's a cylinder, and the formula is pi r squared times length, times 64 pounds per cubic foot) before deciding to jump, you'd be tiger food every time. Maybe Spock could tell you how heavy the tiger is, while both of you are running for your lives. I still haven't got it figured out (my calculator just told me: 1,809 pounds. Oh, well; real tigers weigh 500-700 pounds). Our logical, intellectual life is only useful when we have leisure to take advantage of it. Anchoring may be a useful trait when choosing which of two cliffs to jump off; you'll always choose the lower one, and you'll be right. It just doesn't translate well into a retail setting, unless you are the retailer!

It takes a bit of humility to read this book, but that won't hurt. Many of our follies are undergirded by the tendency we all have to overestimate ourselves. Something like 80% of drivers say they drive better than average. The definition of average guarantees that only 50% (minus 1) can be above average. And before someone asks, "That's only true if 'average' means 'median'!", I agree. Because that is what I meant. Driving skill can be ranked but not quantitatively scored, so median is the only average that is possible.

I learned a few things that may help me. Chances are I'll forget most of it, though. I pride myself on a good memory, but 48 informal fallacies? Beyond me. It was a fun read, though.

Brainy

kw: observations, intelligence

This chart, available from many sources, shows the brain-versus-body size and the Encephalization Quotent (EQ) for a number vertebrates. These are averages, so you can imagine each dot to represent a small cloud of points. The bold line marked "Higher vertebrates" shows the trend for mammals and birds, while the one for "Lower vertebrates shows the trend for reptiles, amphibians and fish.

The bold lines have a slope of 2/3 (in logarithmic terms), and are the scaling axes for EQ as a rough measure of relative intelligence. The upper line is drawn such that a 4 kg house cat would fall on the line, and is defined as the EQ = 1.0 line. The lower line is the EQ = 0.1 line. By proportion, you can see that humans, whose dot is near top center, are somewhat less than EQ = 10, and in fact if you use the formula shown along the line, an average person would have an EQ of about 7 to 7.5.

Just to the right of that point, "Porpoise" represents the bottlenose dolphin, which has an EQ in the range 4-5, making it the "brainiest" sea creature. I read somewhere that an octopus, I don't know which species, has an EQ around 3 or 4, but their brain is so much differently organized than the vertebrate brain—for one thing, it is ring shaped—that the measure may not be a meaningful comparison.

I don't know whether EQ truly correlates well with "intelligence". For one thing, we don't yet have a good definition of intelligence. But it is known to correlate well with the amount of time young animals spend playing. Humans of all ages, dolphins of all ages, and the young of elephants, octopi, dogs, cats (and many adult dogs and cats), and many kinds of birds have been observed playing. Even mice play. I haven't read whether fish or reptiles play. Maybe somewhere around EQ = 0.25 is a threshold below which play is not found.

Apparently, somewhere near EQ = 0.05 is a threshold of the amount of brain needed to keep a vertebrate body from dying outright. A bigger brain than this has added abilities to modify the environment (such as by building a nest) and conduct social interactions. The more "extra" brain there is, the more capable such functions become. I suppose. I am still waiting for a more clear answer about the amount of brain, either absolute or relative to body size, needed to form a self-concept and self-consciousness. The early hints are that it is near EQ = 2, where Chimps are found. Once we learn more, lots more, about natural intelligence, I hope we'll also be able to produce machine intelligences, or at least to know what it would take (IBM's Watson is still a far cry from a robust machine intelligence, and it is a multi-million-dollar stack of computers and databases the size of a bedroom).

I suspect the threshold of language is near EQ = 3, maybe less, so our efforts to communicate with dolphins (so far with poor success) make it clear that, if SETI ever does detect the radio broadcasts of space alien civilizations, we'll be in for a long, long learning process to figure out how to understand what they are saying.

When feelings and thoughts collide

kw: book reviews, nonfiction, sociology, intelligence

In Hawaiian and classical Egyptian cultures, a king was required to marry his sister. It was unheard-of for either party of the couple to disagree. In contemporary Western culture, a highly distasteful act is said to be "like kissing your sister." There's the old proverb that "one man's meat is another man's poison", and it clearly refers to more than just foodstuffs. How are we to understand the visceral reactions people have to one or the other circumstance? I am sure old Pharaoh would be appalled at our attitude towards his marriage. Some moral choices really are culturally derived. Others less so. I won't say more than that, however.

On a milder note, the next time you watch a baseball game, pay attention to the way the fielders go after a high fly ball. Many times, one will simply trot over, almost unconcernedly, and catch the fly without breaking stride. Other times, one will run desperately for the back fence, only to watch the ball fly just overhead. How does the first player know where the ball will land, and how does the second know it is beyond catching?

It turns out that fielders in baseball take advantage of a principle of calculus that needs no calculation: If you gauge your running speed so that the angle to the ball remains constant, you'll get to the ball just when it comes within reach. There is a way to prove this with calculus, but we know the ball players aren't using any calculus, and very likely none of them know of the intersecting tangent principle anyway.

The first player sees, after the ball stops climbing, that it appears to be drifting in one direction. He trots or runs until the drift apparently stops. Then he knows he can catch the ball. The second player sees that, no matter how fast he goes, the ball's angle keeps getting ahead of him. He hopes he'll be able to jump for it at the wall. The faster that angle is changing, the less chance he has.

Nobody taught this to the players. They learned it by experience. They don't think about it. If you ask one, they usually can't say what they are doing. They "just know" how fast to run to intercept the ball. They don't have to know where the ball is going to land. They just know, if they keep the angle constant, the ball itself will lead them to its landing point…if they think about it at all.

Life is full of heuristics like this. Without them we could hardly function. We seem to be hardwired for grammar, for example. How else to explain that we all use grammar with a measure of correctness (regional and dialect differences aside) that no machine can match, but very few of us could explain the "rules" of the grammar we use with such near-perfection all day long.

Psychologist Gerd Gigerenzer has written Gut Feelings: The Intelligence of the Unconscious, in which he pries into a number of these heuristics. The book rewards careful reading (which is why this post is a couple days late). He investigates many of the choices we make, and how they are made. Consider the Chain Store Paradox.

You own a chain of widget shops. A competitor plans to open similar shops, one by one, in all the areas in which your shops are found. How do you respond?

A logical process called backward chaining reasons thus:

I have twenty shops. If the competitor sets up nineteen shops, then it makes little difference if a twentieth is set up, so I ought to share the trade. By the same token, the difference between the eighteenth and nineteenth shops is similarly negligible. And so on, to the first shop. Therefore it makes little sense to engage in a price war.

However, is this what happens? No, you would most likely try, with the first opening, to drive it back out of existence, even selling your widgets at a loss for a time. The simpler evaluation is, "Every competitive shop will cut into my business, so I need to protect my business." Period.

The message of this book? Simple evaluations often outperform more complex ones. For example, suppose you are asked, which is the larger city, Omaha or Philadelphia? How many people know that Philadelphia is four times as large as Omaha? "Only four?" some may ask. Why? Because we hear of Philadelphia so much the more. To most people, Omaha is where "Prudential of Omaha" is from, and that's about it. Go overseas, and you'd be hard-pressed to find people who've heard of both cities; most know Philadephia, very few know Omaha. And the heuristic does work out, that the more well-known a city is, the larger it tends to be.

If you give people a test, to tell among twenty pairs of city names, which is largest, those who recognize about half the names do the best. Those who don't recognize any of the names just guess. They don't have any knowledge to help them. Those who are very familiar with geography and know nearly all the names, but few details, also do badly, because the "recognition heuristic" doesn't help them. But when you know only about half the cities on a list, it is likely that most of the ones you know are larger than the ones you don't know. This turns out to be true about 80% of the time.

Have you ever flipped a coin to decide between two seemingly identical choices, only to find you "didn't like" the coin's result? The coin forced your unconscious intelligence to assert itself. Our mind is able to make sense of incomplete data. It excels at that. This is where our hunches come from. As the author shows, many of our hunches are smarter than more complex ways of deciding. And that is a good thing, because doctors are so pressed for time, they have to diagnose your condition based on rather vague clues; some of us are still living because many of them get quite good at such "hunch" diagnoses. When this or that guru recommends, "Trust your gut", it is often good advice.