JDN 2456305 EDT 15:30

Within a few years, a decade or two at most, we could easily have a fundamentally new economy, built not around the allocation of limited resources or the maximization of profit, but the satisfaction of human needs. And yet, we may not... for our cultural norms are standing in the way.

Norms, that, on their own, seem reasonable enough: "Everyone should receive in proportion to what they produce." "The best person to make something should be the one who makes it." "If possible, everyone should have their basic needs met." Yet the three of these together, combined with some basic facts about productivity, yield... impossibility.

We are now at the historical epoch in which one person can literally produce 1,000 times as much as another person. How? By inventing a robot that performs a task that previously took 1,000 workers to do.

When this happens (and it does rather often), the "best-person" principle means that the robots get used and the workers get laid off. The "proportion" principle means that the inventor gets paid for the robots and the workers receive nothing. And then the "basic needs" principle? Well, it pretty much gets ignored.

In reality, there are serious problems in our intellectual property system and corporate regulations which mean that it's usually not the actual inventor of a new technology who receives the income that technology generates. But even if it were, would that really be what we want? Given the choice between two worlds, one in which a thousand workers live middle class, and a second in which one inventor is a billionaire and those thousand workers starve, would we really prefer the latter?
In fact, there is a third option (a Third Option!): A world in which the thousand workers live middle-class and the inventor is a millionaire instead of a billionaire. How is this possible? Well, the robots really are more efficient, so there's more stuff to go around. We can give most of the surplus to the inventor, and then everyone else gets what they already had plus a little bit extra. The beautiful part is that this is incentive-compatible; everyone is better off with the robots installed, so there's no reason for the workers to object or the inventor not to do it.

The problem, of course, is that this is not what a competitive capitalist market would produce. Capitalism starts with world one, and leads to world two. World three is never even considered.

And the problem, I think, is a very deep one: It is competition. It is the notion that economics (or life in general!) is about competing, where some people win and other people lose. As long as humanity continues to live in those terms, the world will never be at peace.

We must re-evaluate our entire concept of what economics is about, to one of collaboration. How can we help each other? What does each person have to contribute? How can we make the best of the resources available to us? We are all a family (literally); can't we live like one?

Is this socialist? I suppose it is. Maybe socialism is exactly what the world needs.

JDN 2456304 EDT 18:20

It's long been a mystery in economic decision theory: Why do people play the lottery? Aren't they risk-averse? They seem risk-averse when they buy insurance, but then when they play the lottery they seem massively risk-seeking. So what's going on?
Well, I asked my mother why she plays the lottery, and she explained it thus: "I'd feel terrible if my numbers came up and I hadn't played." This reminded me of a principle in decision theory that doesn't get a lot of attention: Regret minimization.

"Regret" in this formal sense is the difference between what you got and what you could have gotten with your best-case alternative. Playing the lottery has, say, a 99.999,999,999% chance of losing $1 and a 0.000,000,001% chance of winning $100,000,000. The expectation value for this high-risk option is negative: -$0.90. But the regret is minimized if you play, because that's a regret of $1 instead of a regret of $100,000,000.

The beautiful part is, this also works for insurance! If you buy insurance, you pay perhaps $10,000 a year in order to avoid a 5% chance of losing $100,000. Once again the expectation value is negative, but this time it's a low-risk option: -$5,000. Yes, but the regret is only $10,000, whereas if you didn't buy insurance your regret might be $100,000.

Regret minimization has one very big advantage: It's independent of probability. So, since human beings are bad at estimating probabilities, our brains respond by avoiding the problem entirely, using heuristics that aren't dependent on probability at all.

You can still be Dutch-booked this way (indeed, a lottery is arguably exactly that), but it's a lot easier on your computational resources than actually trying to compute a full Bayesian risk assessment.

Therefore, we aren't properly considered risk-seeking or risk-averse, and prospect theory doesn't work either (because it says we should be risk-averse for gains but risk-seeking for losses, and lotteries are risk-seeking gains and insurance is risk-averse loss!); instead, we are regret minimizers.

JDN 2456296 EDT 12:01.

"I know it when I see it," went the famous ruling, by Supreme Court Justice Potter Stewart. Well, that's great... shall we show you every possible example? You'll be the arbiter? The backlog shall be astronomical.

Clearly if we are to make rules about "porn", we need a better definition than that.

One option, I suppose, would be to not regulate at all, just let people use the word "porn" to mean whatever they want and leave the government out of it. That does raise a problem however: Do we really want 7-year-olds to be given access to everything that's on the Internet? (I suppose one could argue that they already do; but we at least try to prevent that.)

Instead, I think we should ask: What are we worried about porn doing to us?

For some people, the mere fact that children would be exposed to sexual content at all is horrifying; I'm going to set that aside, because frankly it's ridiculous. Humans are sexual animals, and kids are curious about sexuality from a young age. There is such a thing as age-appropriate sexual content.

In fact, the kind of sexual content that is in my books right now is toned down relative to what I wrote when I was a teenager, and yet I'm told that my books are inappropriate for young adults. Apparently what young adults write is inappropriate for them to read?

I think when I worry about porn damaging impressionable minds, my fear is that it will distort sexuality, give people wrong ideas about how sex works and what it is like. Some of this is inevitable: You can't smell the odors or feel the wetness on a computer screen. But some of it is not: Why do all porn actors shave their pubic hair off? Why are they so short, so thin, and with such enormous breasts and penises? Why do they use such awkward, bizarre positions?

I suppose it's too much to ask that ten-minute videos be honest about the psychological and emotional experience of sexuality. But what about two-hour-long films, or 50,000-word novels? Is it too much to expect that they be realistic about the joys and harms of sex? It seems to me that if we are going to define "porn" as something problematic, it would be this:

Pornography is content that distorts sexuality. At best, it idealizes it; at worst, it makes sexuality into something degrading and commodified.

I feel similarly about violent content: A historical drama on WW2 or the Civil War is bound to be violent; but done properly it could be honest about the violence, and show just how terrible and horrific it really is. I'm much more disturbed by something like Tom & Jerry cartoons, where violence is depicted as whimsical, harmless fun. A game like Call of Duty is actually about the horrors of war (when you die, you get a quote about how awful war is!), and Halo seems relatively harmless because it takes place in a futuristic world and you kill aliens. But Grand Theft Auto? It's just pure nihilistic, psychopathic violence. You kill people because you can; the proper authorities are your enemy. If I had my druthers, Grand Theft Auto would be much more regulated than Call of Duty.

It's interesting though: By this definition, much of what is currently not considered "porn" should be! Supermodels, makeup ads, and teen pop stars all distort sexuality in obvious ways. Meanwhile, videos of a young woman masturbating? They don't seem all that distortionary, and thus might not even be classified as "porn"!

This result might seem counter-intuitive, but it actually makes a lot of sense in terms of the real social impact of such media. Children who see airbrushed pictures of supermodels are subtly influenced toward unhealthy body image, which we know can lead to depression and eating disorders. Children who see videos of young women masturbating would be advanced in normal psychosexual development, and might well learn techniques for pleasuring themselves or their partners later on.

This refined definition of "pornography" could also lead to some compromise between pro-sex feminists and anti-porn feminists; if depictions of sexuality are not automatically considered pornographic, then we can be pro-sex while still being anti-porn. Conversely, if pornography is defined as that which distorts our sexuality, it makes sense that we would find it problematic.

And make no mistake: Most of what we currently call "porn" would still be called "porn", because it does distort our sexuality, sometimes in terrible ways. Mainstream porn's standards for what our bodies should look like and what we should do in bed are bizarre; given the prevalence of unprotected multi-partner sex in porn, they are outright dangerous. Some of the material that anti-porn feminists object to is genuinely objectionable. (Then again, sometimes they get pretty off-the-wall: I've heard it argued that gay porn objectifies women... by not including them. Also anti-porn feminists are notoriously unwilling to listen to complaints that mainstream porn objectifies men, even though it obviously does.)

Should this be regulated? Yes, I think it should, though precisely how I'm not sure. There is a very delicate balance to be found here, between protecting impressionable minds and preserving freedom of speech. The LAPS test helps to some extent: Content with "significant literary, artistic, political, or scientificmerit" is already excluded from being regulated as pornography. The problem is, who decides that? Political and scientific merit are at least reasonably well-defined, but literary and artistic merit are notoriously indefinable. Also, I'm not sure we really follow this for political merit; I'd imagine we would probably censor a cartoon of the IMF graphically raping Africa as Dominque Strauss-Kahn allegedly raped an African woman, even though the political implications are ludicrously obvious (and not altogether inaccurate either). Come to think of it, we're actually better about protecting literary merit: Lolita and Ulysses were both protected by Supreme Court rulings.

JDN 2456292 EDT 16:33.

The Kama Sutra is a great book, and I really mean "great" in the sense of Hamlet or Iliad or On the Origin of Species. But it is also deeply flawed, and like the Bible its flaws have been preserved over the centuries in the name of religious reverence. Unlike the Bible, the Kama Sutra also has a very open attitude about sexuality (which is how most of us in the West know it), which has also made our puritanical culture wary of discussing it at all, let alone to critique it seriously on its own terms.

At its core, the Kama Sutra is a truly wonderful notion: Sexuality as a means to spiritual transcendence. It's about elevating sexuality, expanding its potential. The Kama Sutra is a sex manual, but it is a sex manual for mystics and philosophers.

But some of the things it gets wrong... they're awful. Basic facts about the functionality of orgasm, particularly female orgasm. It contains a long section debating whether female ejaculation exists (hint: it does, period), completely confusing it with orgasm (it's not even the same thing in males, for goodness' sake). It reasons a priori, much like Plato and Aristotle, which means it just plain gets a lot of things wrong about physiology.

Also, while it rates levels of desire as "small, middling, intense" (pretty straightforward), it has a really bizarre categorization for size, "stallion/elephant, bull/mare, hare/doe". Now that just makes no fucking sense, pardon the pun. Stallions evolved to have sex with mares; their genitals correspond, you idiots. Also, elephants and hares come in both sexes. If you want to use animal comparisons (which are obviously exaggerated), here's a much more logical system: "stallion/mare, bull/cow, dog/vixen". Or, you know, you could just go with, "small, medium, large". In fact, there are at least two independent dimensions of size involved here (three geometrically, but the transverse dimensions tend to be strongly correlated), where girth is often much more important than length. And then there are shape considerations as well.

I'll tell you what I do like about the Kama Sutra's size system however: It focuses on compatibility. In our own society we seem to think that a huge penis and a tight vagina are objectively the best; but see, those don't actually fit well together--in fact, sometimes they literally do not fit at all. Men with small penises shouldn't be made to feel bad; they should instead find women with small vaginas. Likewise, women with large vaginas shouldn't be put down, they should find men with large penises. The Kama Sutra gets this right, and while it describes the combination of large penis with small vagina as "high union", it explains that "high union" is not nearly as good as "equal union" in which the two fit comfortably. Likewise, it doesn't tell men to "last longer" or women to "come faster" (as so many pills falsely advertise today); it says that different people take different lengths of time to reach orgasm, and you should look for someone whose timing is compatible with yours. To be perfectly honest, I sometimes wish I didn't take as long as I do; I literally have to schedule around it, and can't just do the five-minute quickie some people can. Soon we will probably have medications that can actually modulate orgasm timing, but it's not obvious to me what the best time would be, other than aligning with your partner's.

The Kama Sutra also doesn't talk about any of the risks of sex, either psychological (e.g. heartbreak) or biological (e.g. infection). It presents an idealized vision of sex as always perfect and beautiful, and thus doesn't prepare the reader for the very real risks and disappointments that sex can involve. Admittedly, I prefer this to the way most sex education classes are presented, in which sex is depicted entirely in terms of risks, harms, and dangers. But still, sex does carry real risks, and we should be honest about them.

The Kama Sutra is most famous for its sexual positions, and some of what it recommends are really bizarre, difficult, and I must imagine not terribly comfortable or pleasurable. Then again, it's written by the same class of mystics who brought us yoga, so maybe they're writing for an audience well trained in bodily flexibility. For those of us who don't bend in those directions, it would be nice to have a spiritual sex manual that focuses on more traditional positions, like missionary, cowgirl, and doggystyle.

Also useful would be an integration of queer sexuality: gay sex, lesbian sex, polyamory, pegging, kink, and so on. Teach people that queer sexuality is really not so queer, that most people are "deviant" in some ways and most of this "deviance" is utterly harmless. Largely gone are the days when straight people literally don't know how to have sex (Kinsey wrote about a few such couples he interviewed), but gay and especially lesbian couples still have this problem sometimes. It doesn't help when sex education classes barely even mention homosexuality and provide detailed information about the vas deferens and fallopian tubes without even the briefest mention of how to have anal or oral sex. Instead of an education about sexuality, we get one about reproductive physiology. This would be rather like having a culinary arts class and spending the whole time talking about the function of the intestinal tract.

I think the reason we don't talk about these things is that we fear teenagers will have sex more if we tell them how to do it properly. There's actually no evidence to support this (rates of teenage sex do not increase when comprehensive sex education is introduced [citation]), but even if it were true... so what? Suppose it really were the case that when we teach teenagers how to make sex safer and more fun, they have more sex, because it's safer and more fun. What would be bad about that?

To me, that feels like saying, "If you give kids helmets, they'll ride bicycles more." Yes, that may be true; and the problem is? I think the unstated assumption, the implicature we're all supposed to read between the lines, is sex is bad; so obviously more sex would be a bad thing. But I simply don't see it that way! More infection is certainly bad, and more pregnancy could be bad (especially unwanted pregnancy), and even more heartbreak could be bad... but more sex? If you have reduced the risks, you've reduced the risks, and I thought that was the whole point.

This is actually how I feel about risk compensation in general; it's often argued that safety features don't really do anything, because people provided with safety features often take more risks [citation]. But they take more risks and get more benefits, which is why they do it. Install airbags, and we drive faster; but that means we get where we are going sooner. (And if you don't see why arriving sooner is worth risking injury or death... why don't you walk everywhere you go? Or just stay home for that matter?) Moral hazard is a real danger--when my actions incur risk to you, we have a problem--but risk compensation just means that people are willing to take risks to get benefits, and as well they should be. I don't mean to imply that all risks people actually take are rational; far from this, plenty obviously aren't, like smoking, football, binge drinking, and indeed reckless driving. But some risks are rational, and risk compensation is not inherently a bad thing.

There are some modern books that do a fairly good job of explaining sexuality honestly and usefully, like The Joy of Sex, The Underground Guide to Teenage Sexuality, The Guide to Getting it On, and Sex: A Book for Teens. But none are comprehensive, and I can't think of any text that does a good job of taking seriously the transcendent component of sexual psychology without falling into ludicrous mysticism. (And indeed, the Kama Sutra falls headlong into ludicrous mysticism.)

In fact, it's rare to find a book that talks about spiritual transcendence at all without this sort of lapse into ridiculousness, but there are a few: Spinoza, Einstein, Bertrand Russell, David Bohm, Carl Sagan, and lately Sam Harris. Yet I've seen hardly any yet that talk about sexuality explicitly in those terms, probably because they are afraid of puritanical backlash. This is how religion controls us: It takes away our real joy and then offers us imaginary substitutes. It's a brilliant scam, really.

Perhaps I should write this new sex manual, a Kama Sutra for the twenty-first century. Of course, I still have two novels in progress, a finished nonfiction book seeking publication, and another nonfiction book in draft stage, not to mention short stories...

JDN 2456291 EDT 12:13.

I realized today why so many people have trouble understanding concepts in cognitive science that to me feel so elegant and intuitive. I had been assuming that other people think much the same way I do (Mind Projection Fallacy), and I hadn't been correcting for my highly unusual personality traits.

Specifically, I am an empathizer-systematizer. Most people are either empathizers or systematizers, not both. Empathizers conceptualize the world in terms of conscious beings with thoughts and feelings; they relate to others on a personal level, try to view the world from their perspective. Systematizers conceptualize the world in terms of physical phenomena, obeying natural laws, organized into structures made of parts. While most people have some degree of each trait, the majority of people are much better at one than the other. Also, statistically women are more likely to be empathizers and men systematizers, but it's not as large a difference as stereotypes would have you believe (Marie Curie and Rosalind Franklin were classic systematizers, and Martin Luther King and Mahatma Gandhi were textbook empathizers).

You need to be at least part systematizer to be any good at science. If you're a pure empathizer, you'd make a good social worker or psychotherapist, but scientific research is always going to be hard for you. If you're an extreme systematizer but a very low empathizer, you can end up with a "mad scientist" sort of attitude, even bordering on psychopathic--B.F. Skinner is a great example of this. The things he did to animals were unconscionably cruel, but he didn't think of them as animals; he thought of them as complex systems of interacting parts--so he took them apart and studied the pieces!

I am a bizarre case, an empathizer-systematizer, someone who is good at seeing the world from both of these perspectives simultaneously. And this is what you need to be able to do in order to make sense of cognitive science. You need to be able to look at a brain and understand the way it can be broken down into lobes, sulci, neural clusters, neurons, synapses; but you also need to understand how that brain houses a mind, with thoughts, beliefs, memories, hopes, feelings, desires. You need to be able to appreciate the marvelous fact that minds are made of parts.

Steven Pinker is also an empathizer-systematizer, which is why he makes a good cognitive scientist. Noam Chomsky is as well, though he lets his empathy cloud his reason when it comes to political issues. Steven Pinker's politics make a great deal more sense, though he occasionally ventures a bit too right economically. Still, center-right libertarian is far more sensible than far-left anarcho-syndicalist. In any case, I think you'll find that most of the top cognitive scientists are all empathizer-systematizers.

Also, my boyfriend is an empathizer-systematizer, which is probably why we get along so well. Kittens and tabletop RPGs are two of his favorite things. No doubt it also helps that he is a sensitive, highly intelligent, introverted, and the nerdiest person I've ever met.

I know a lot of systematizers who aren't empathizers, and conversations with them about cognitive science often lead to them making really bizarre greedy-reductionist claims. They can see the parts, but they can't see the whole! "Cognitive science will be ultimately reduced to physics." No, except in the sense that thermodynamics is reduced to statistical mechanics, which is pretty much already the case. (I don't know of any serious cognitive scientist who thinks that the brain operates on non-physical principles! Even Penrose, who is on the fringe, thinks the brain operates by physics we don't yet understand.) "There is no fundamental difference between a human being and an asteroid." Funny, I don't seem to be able to have this conversation with an asteroid. (I suppose it could hinge on what we mean by fundamental, but I never disputed that we're made of atoms. I merely point out that whole humans and whole asteroids can and should be treated differently.)

It's not just personality of course. Computer scientists tend to understand better what I'm getting at, even though most of them are systematizers and not empathizers. I think it's because computers and minds really are so similar that the two can be used to understand one another. The sorts of things that physicists and biologists say to me about cognitive science would be translated into computer science as something like, "It's all made of zeroes and ones!" or "It's all electrical circuits!" Yeah, so what? That doesn't help me optimize heuristics for the traveling-salesman problem.

Economists are better about it too, even though economics clearly is in desperate need of more empathizers. (Michigan's own Frank Thompson is one obvious exception, a clear empathizer-systematizer; he's a big part of why I got into economics in the first place.) The condescension of economists comes from a different direction; they think psychology is "soft", "fuzzy", too much talk of emotions and personalities instead of rigorously defined behavior patterns. But they would never claim that it can be reduced entirely to quantum physics; they know how hard it is to model the behavior of a small number of rational individuals, much less a large number of irrational ones.

But personality clearly does matter. Even though our training is quite similar (classes across the hall from each other), social psychologists are almost all empathizers, and they tend to view us cognitive scientists with a sense of awe. They struggle with t-tests and ANOVA as we solve differential equations and compute Bayesian neural networks. They are probably what economists are thinking of when they think of psychologists (though even then, I think the economists underestimate the value of social psychology); but the social/cognitive divide runs much deeper than people outside psychology realize. And only we psychologists understand why: It has to do with the personality types of people who choose these different specializations.

What I need to do now is find a way to explain cognitive science to people who don't have the requisite personality, find a way to express the deep insights in terms that pure empathizers or pure systematizers can understand. Thus far, I've found empathizers a bit easier to work with; while they clearly don't understand the details of what I'm saying, they rarely dismiss it outright and often remark that it sounds fascinating. It's the systematizers who are most frustrating; they simply dismiss the empathic perspective as naive and fuzzy-headed, so they feel that once they've understood each part by itself, obviously the whole will just magically fall into place. This is how we got into the pathetic muddle of behaviorism, eliminativism, and epiphenomenalism. From the other end, pure empathizers gave us mysterianism.

To see just how stupid these theories really are, allow me to make an analogy to physics. We did the two-slit experiment, got this crazy weird result that we couldn't understand, and here's how they reacted.

Behaviorist: Ignore that, we don't understand it so it must not be important.

Eliminativist: That obviously didn't happen, you're just naively accepting the folk notions.

Epiphenomenalist: Well, maybe it happened, but don't worry, it doesn't affect anything. It's just sort of added on to physics, and doesn't do anything, I'm sure.

Mysterian: You see? Physics is a failure! We will never understand the mysteries of the atom!

I don't think I'm going out on a very long limb when I say that we would not have invented lasers or microprocessors from any of those lines of research. And likewise, it will be cognitivists, not epiphenomenalists or mysterians, who solve the problems of AI.

Because computer scientists are systematizers but understand better, I'm thinking I'll need to express what I'm saying in computational terms, using words like "hardware" and "software" and "algorithm"; but even when I do this, systematizers often still have a dismissive attitude. Maybe it would help if they actually studied some computer science, and learned just how absurdly difficult some computational problems really are. Many physicists and biologists seem to think that computers can just improve in complexity indefinitely and become capable of solving any imaginable problem by brute force, thus rendering algorithms and heuristics irrelevant. If they actually read some computer science, they'd realize that the game of Go is too complex to be solved by brute force by a Planck-scale computer the size of the Earth in a thousand years.

Here's the calculation, if you don't believe me. There are 19x19 spaces on a Go board, each of which can be empty or filled by one of two colors. That's 3^(19x19) possible states, which is 1.74e172. The Earth is approximately sphere of radius 1.3e4 m, volume 4pi/3*(1.3e4 m)^3 = 9.2e12 m^3. A Planck length is 1.616e-35 m, so there are 2.37e104 Planck volumes in a cubic meter. That means our computer would have 2.18e117 components, each performing one calculation per Planck time (5.39e-44 seconds), which means our computer calculates at 4.05e160 flops, which is mind-bogglingly fast... and yet, our total calculation would still take 4.3e11 seconds, which is over 13,000 years. I remind you, that is for the game of Go, and it is a very generous estimate of the fastest possible computer anyone could build out of our planet. This is why Go programs do not use lookup tables in the form "if board X, do move Y"; it would take eons just to sort through the table, which by the way would have to be a trillion times the size of the Earth even if each Planck volume could store 1 bit. With current technology for storage and computation, the hard drive would be the size of the galaxy and the processing would take more time than the universe has before heat death. Real Go programs use heuristics, and actually they're not all that good; I can beat most Go programs, and I'm not even good enough to play in most amateur leagues. If you can find better heuristics, good enough to beat high-level pros, you could find fame and possibly fortune creating the "Deep Blue" of Go.

And you're telling me that human behavior is going to be understood by brute-force computation? No, we're going to use heuristics. And those heuristics are going to involve high-level systems, like beliefs, desires, emotions, and intentions.

Now, you might say, this is an unsatisfying philosophical result: It sounds like I'm saying we are machines, but we have to pretend we're not because we'll never make sense of anything that way. No, that's not what I'm saying. We're not pretending anything. We are machines--we are machines that think. The empathic perspective works as a model because it accurately reflects the world; its heuristics are not chosen arbitrarily, but rather necessitated by the structure of the phenomenon. Biologists of all people should understand this: Is there such a thing as a giraffe, or are there merely atoms? No, the giraffe is a real thing; there is a meaningful (if slightly fuzzy-edged) boundary between the giraffe and the surrounding world. We carve the world, but we carve it at its joints.

This is the empathizer-systematizer part, the part I don't know how to explain: Both perspectives describe the same real phenomenon. We are machines, and we do have feelings--and the marvelous thing is that machines can have feelings. It is not either-or; it is both-and.

Hinduism has a parable about this, which isn't bad: Two blind men are trying to understand an elephant. One feels the trunk, and says it is soft, bending, like a snake; the other feels the legs, and says it is firm, rigid, like a tree. Who is right and who is wrong? Both? Neither? The whole elephant encompasses more than either part. If you truly wish to understand the elephant, you must see it as a unified whole.

JDN 2456285 EDT 22:47.

Many reviewers praised The Psychopath Test for its irreverence; personally, I think that's its greatest flaw. The book is certainly engaging, and a very quick read; Ronson writes like a journalist, so his style should be accessible to just about anyone. He makes a lot of cute jokes, some of which are quite funny, and they certainly do lighten the mood tremendously.

And yet... I'm not sure I wanted the mood lightened. The topic is psychopaths, after all; and increasingly we are finding that a long list of terrible things can be attributed almost entirely to the 1% of people who are psychopaths: rape, murder, car theft, bank robbery, Ponzi schemes, insider trading, tyranny, genocide, the list goes on. At the end of the book Ronson takes a step back and wonders if we are over-diagnosing or over-stigmatizing psychopaths; yet all the evidence I've seen says quite the opposite: We should be rooting them out and locking them away until we can find a way to cure them. The cost of such a program would be great, but the benefit is almost unfathomable: Imagine a world without rape. Imagine an end to war, to tyranny, to genocide. These things might literally be possible if we could eliminate psychopathy.

Maybe some mental illnesses are over-diagnosed: It can be argued that "autism spectrum disorders" are too broad a category, for example, since they range from the brilliant but socially awkward (like myself, or Bill Gates) to children who can't speak and bang their heads against walls. ADHD is a real condition, but some of the kids diagnosed with it might just really be hyperactive kids who need more recess. Bipolar disorder can be debilitating to those who really have it, but lately we've been slapping it onto children for the weakest of reasons.

But psychopathy? No, it is not over-diagnosed. It is under-diagnosed; millions of psychopaths are lawyers, bankers, executives, politicians, and military commanders, and their reckless and callous actions endanger us all.

Ronson definitely knows how to write, and sounds like a very interesting character himself. The book reads quickly and is difficult to put down. It feels like he's sitting with you in a bar, telling stories of all the crazy things he's done in his life.

Perhaps the best part of The Psychopath Test is in fact the psychopath test itself; Ronson presents a simplified form of the actual Hare checklist that is used to diagnose psychopathy. You may be tempted to apply the checklist to people you know, but you should be careful doing so; even with appropriate training it's difficult to make the necessary judgment. That said, if you know someone who fits a lot of the criteria, it makes good sense for you to be cautious around them. I know a guy from high school who fits so many of the criteria that honestly, I don't think I want to have anything to do with him ever again.

The book is worth reading, not least because you can finish it in a few days. But don't believe Ronson when he suggests we should leave more room for madness: Psychopathy is one form of madness we certainly cannot leave more room for.

JDN 2456283 EDT 17:23.

It's taken me some time to realize this, for a few reasons. My high level of empathy makes me want to see the good in everyone, my liberal sensibilities make me look first to social structures and systems of policy to explain behavior, and my atheism makes me dubious of beings of pure evil like demons or Satan.

Yet, there are bad people, people who are, innately, probably genetically, wired differently from other human beings. They are almost not human beings, in fact; they lack the traits that made human civilization possible. They are in some sense my opposite: Where I have very high empathy, they have virtually none. Where I struggle with self-confidence, they are convinced of their own superiority. Where I feel guilty for saying the white lies that social norms expect, they deceive and manipulate without hesitation. We call them psychopaths.

Not all the bad things in the world are the result of psychopaths: There are crimes of passion, honest mistakes, bad policies, and false beliefs, all of which do plenty of damage. Yet, the really terrible things, the truly monstrous evils, are almost always committed by psychopaths. Organized crime, corruption, serial murder, rape, tyranny, genocide--psychopaths, almost without exception. Indeed, while religion makes good people do bad things, religions are almost always founded by psychopaths. This should be obvious in the case of L. Ron Hubbard and Joseph Smith; but I would hazard a guess for Muhammad and Moses as well. The only one I hesitate about is Jesus, who clearly showed empathy; he seems more like he had a very severe case of psychotic bipolar disorder: Liar, Lunatic, Lord. Then again, Christianity didn't really take off until Constantine, who was definitely a psychopath. Increasingly, we are finding that the instability of our financial system can be traced in large part to a small number of psychopaths manipulating the system in their favor at our expense. World War 2, the worst conflict in human history, can basically be traced to two psychopaths, Adolf Hitler and Josef Stalin.

In fact, psychopaths are all around us; they comprise about 1% of the population, which means that you probably know a few, and there are 3 million in the US alone. A lot of people, even liberals, were saying that it's no big deal that Mitt Romney used to bully other children and put a dog on the roof of his car; but it is a big deal: Juvenile behavior problems and cruelty to animals are classic signs of psychopathy. Do you want a psychopath in charge of the American government (and military, and nuclear arsenal)? Because Mitt Romney fits most of the criteria for a high-functioning white-collar psychopath. So does Dick Cheney, and he almost single-handedly got us into the Iraq war, which killed almost a million people; as far as I can tell he did so primarily to raise the stock price of Halliburton (which he of course owns many shares of).

Lest you think I just tar people I don't like with the charge of psychopathy, I am fairly certain that George W. Bush is not a psychopath; he doesn't fit most of the criteria. Nixon probably was, but Reagan probably wasn't. I don't like the policies Bush and Reagan enacted, but I think they genuinely believing they were making America better. They weren't simply acting in their own narrow self-interest the way psychopaths would. Actually, staunch opposition to abortion and gay marriage, while awful in its own way, is very unlikely for a psychopath, because these issues require emotional reactions and consideration of issues beyond oneself. Notice how Dick Cheney flip-flopped on gay marriage when he found out his daughter is a lesbian; this might seem like a sign of compassion, but it makes just as much sense as a strategic decision on an issue he never really cared about. Someone who feels the sense of disgust and repugnance that makes them want to restrict homosexuality has a kind of emotional repertoire that psychopaths don't. (On the other hand, it's possible to be disgusted by something and still not think it's the government's business: This is how I feel about coprophagia, for example.)

When we talk about "mental illness" as a cause of gun massacres, we really should be talking about psychopathy. It is not people with bipolar disorder or autism who commit violent rampages. Maybe there are some people with schizophrenia so severe that their paranoid delusions cause them to become violent (as you might too, if you believed you were surrounded by alien invaders or agents of a tyrannical government), but plenty of people with schizophrenia function normally with therapy and medication.

No, it is psychopaths that we don't know what to do with. We know that they commit a disproportionate quantity of violent crime, and we are finding that they are also responsible for a great deal of white-collar crime as well; but we know no treatment for their condition, and often the best we can do is lock them away and never let them out.

Most people are empaths, i.e. we feel empathy and aren't psychopaths. But this means that most people, being empaths, don't really understand how psychopaths work. They say things like, "If it's their broken amygdalas, doesn't that mean it's not their fault?" No, you don't understand. Remember the Basic Fact of Cognitive Science: We are our brains. Of course there's something in their brains that is different--if there weren't, they wouldn't act different. If we could actually alter their brains to make them into empaths, by all means, do so; I've always felt that A Clockwork Orange is a better solution than either allowing rape and murder or locking people away for their whole lives. But right now, we can't; we're stuck with them being broken, dangerous... evil. That amygdala anomaly is the physical manifestation of evil, which of course must exist, because evil exists and the world is made of physical stuff. You should feel hardly more pity for the psychopath than he would feel for you--which is to say, none at all. If it helps you to think of him as a demon, go ahead; for he is as close as one can come in the real world.

JDN 2456280 EDT 17:06.

The Dosadi Experiment was a novel Frank Herbert wrote in the middle of his career, with some Dune books before it (up to Children of Dune) and some after it (God Emperor of Dune and beyond). Actually, come to think of it, it's roughly "the good Dune books" before and "the bad Dune books" after.

It's a strange novel, longer than it needed to be, and with characters who manage to be complex without being particularly interesting or sympathetic. The closest to sympathetic are the co-protagonists (and by the end, they are co- in a way I probably shouldn't give away), McKie, a galactic equivalent of a federal agent who is practically a James Bond stereotype, and Jedrik, a local to the hostile planet of Dosadi who is far more dangerous as an operative. It's actually a sequel, though I didn't realize that when I started it; the previous book is called Whipping Star, and there was also an earlier short story called "The Tactful Saboteur".

By being removed from the Dune canon, the novel gives us a glimpse into the mind of Frank Herbert himself, revealing the themes he felt were important enough to cross worlds. A surprising example is the chairdog, a genetically-engineered lifeform that apparently combines the functions of, well, chairs and dogs, a piece of mobile, living, affectionate furniture. Much less surprising is the ubiquitous betrayal, ruthlessness, and political intrigue; Herbert revels in this stuff, and frankly it gets so intense in The Dosadi Experiment as to be outright annoying. You never know anyone's real agenda, even the co-protagonists', until the very end (and even then...).

Particularly bizarre are his depictions of women and gay people, which I had hoped would be confined to Dune, but alas, I was wrong. In the ConSentient universe as well, women are superior beings to which men must bow down (this is what ruined the later Dune books), and gay people are depraved. This line was so bizarre I had to stop myself from laughing out loud: "With rare exceptions, primitive Humans of the tribal eras reserved their homosexuals as the ultimate shock troops of desperation. They were the troops of last resort, sent into battle as bersekers who expected, who wanted to die." As far as I can tell, there is absolutely no historical basis for this claim of homosexual berserkers. The Spartans were pederasts... but that's the closest I can get. I'm not really sure what Frank Herbert was smoking here.

Also weird is that once again we have another group, this time another species, who uses their females purely as breeding engines; in Dune it was the Bene Tleilax, here it is the Gowachin. Let me just say... I think Herbert had some issues with women. In all his stories they are either godlike or sub-sentient; the idea of an actually, well, human woman is apparently foreign to him. (Jedrik literally makes no mistakes throughout the entire novel. None. Zero. Everything goes according to her plans. Mary Sue Jedrik should be her name.)

The story had some interesting potential, but I didn't feel like that potential was well applied. The experiment itself involves an entire planet of millions of people that has been isolated for thousands of years in order to evolve in a particular way... and then when you get to the reason why, it feels really anti-climactic and banal. Surely any civilization capable of engineering chairdogs can think of a better way to achieve this objective (which I guess I shouldn't spoil?).

The book also displays more of Herbert's ambivalence toward technology, which actually I sort of like, not because I agree with it, but because it adds a necessary corrective to a general trend among hard SF authors to love technology unconditionally and underestimate its risks. Herbert compensates for Asimov and Roddenberry. It was for this reason that I was very excited to read The Butlerian Jihad, and rather disappointed by its execution by Brian Herbert (Frank's son).

Then again, The Dosadi Experiment does a particularly poor job of discussing the limits of technology, as it grants superiority to the impoverished people who have evolved in a state of violence for thousands of years, rather than the wealthy technocrats who put them there and have ruled them, and presumably advanced in their technology, for those same thousands of years. It would be rather like pitting Rwanda against the United States and expecting Rwanda to win (only many times more so). This was also a bizarre outcome when the Fremen did it in Dune, so once again it seems to be a Herbertism. (Why, oh why, couldn't it be the Spacing Guild versus the Bene Tleilax? That makes so much more sense.)

In all, I was not particularly impressed by The Dosadi Experiment. It's all right I guess; not as good as the early Dune books but better than the later ones. At least there weren't any space-dominatrices conquering men by the power of sex. 

JDN 2456279 EDT 18:57.

The Connecticut school shooting, which you've no doubt heard plenty about already, is but the most recent in a long series of wanton shooting massacres that have been occurring in the United States for just about as long as the United States has existed. Thousands of people have died this way.

And no, this is not inevitable. In fact, most countries don't have this problem at all. The US is an outlier in the First World in terms of its homicide rate and especially gun homicide rate. There are other nations that have higher homicide rates, but none among highly-developed nations. The US homicide rate is about 50 per million per year, while most First World nations have rates nearer to 20 per million and Japan's is as low as 3 per million.

Clearly, there are a lot of things that could be contributing to this. The US has more poverty, poorer health care (especially mental health care), more inequality, and more racism than most other First World nations. But we also have... less gun regulation.

It's not as simple as saying we have more guns: Canada has lots of guns, and in Switzerland every able-bodied male is required to be in the military and own an assault rifle. What we have is instead less regulation of guns, so it's easier for people to get high-powered weapons without background checks, waiting periods, or restrictions. I've literally been in a gas station (near where an ex-boyfriend lived in northern Michigan) that has, on the wall, for sale, a PS-90 submachine gun. It was locked to disallow full-automatic fire (hence the PS instead of just P)... but this should be little comfort, as they are plenty dangerous in semi-automatic and are not too difficult to modify to restore full-automatic capability. In case you didn't know, the P-90 in unlocked form is a military-grade weapon used primarily by special forces. A P-90 is not for hunting Odocoileus virginianus; it is for hunting Homo sapiens.

Ironically, it is illegal for civilians to own tasers in most states, including Michigan. But I could have put this PS-90 on my credit card and brought it home, probably the same day. Is this because tasers are dangerous? They certainly are dangerous... but not nearly as dangerous as guns! Tasers are designed to stun, and in rare cases they can cause permanent damage or death. Guns are designed to kill, and they do it well.

For those who want to use the Second Amendment: Do you really think you can defend against the government with guns? They have aircraft carriers, drones and helicopter gunships. If the US government wants you dead, you're pretty much going to be dead. Your best hope of preventing that comes through civic action, not some fanatic last stand with your gas station P-90. Weapons divide us, and they make civic action harder, not easier, as Firmin Debrabander cogently explains. A Twitter post is worth a thousand bullets, as the Arab Spring clearly showed.

Unfortunately, there are too many gun nuts in America, people who honestly think (despite the clear evidence to the contrary) that owning a gun makes you safer. The fact is, you're more likely to kill yourself, or get involved in an accident, than you are to successfully defend yourself. There's even some evidence that owning a gun makes violent situations escalate--what would have been theft or assault now becomes homicide. Most gun owners have a deeply skewed notion of how crime works.

These are not matters of opinion. They are not complex moral dilemmas. They are objective facts.

But something happened in America, I'm not sure when: Conservatives became immune to facts. And that terrifies me more than anything.

JDN 2456778 EDT 21:26.

Why does the concept of free will lead to such intense disagreement, even among people who otherwise agree? Jerry Coyne mocks hapless bloggers who make bad arguments for free will, but come on, it's easy to make a bad argument for a good position.

In fact, free will seems to be a situation where even once we agree on all the facts, we still disagree intensely on the semantics. Even once all the fact nodes are set, the algorithm demands one more node.

But here's the thing: There are compatibilists (or "requiredists") who recognize this problem, and offer the following solution: "Okay, don't call it 'free will'. Talk about rational volition processes and cognitive decision algorithms." And yet, incompatibilists refuse to accept this solution, and my question is: Why?

I think it's because they really do think there are moral implications of determinism. They have bought into the framework that religious "free will" offered them, and haven't questioned it (or questioned it thoroughly enough). When Sam Harris argues that life without free will would mean no retributive punishment, it's clear he thinks that something very important hinges upon whether we call the faculties we obviously do have "free will".

But there are really only two possibilities here, even logically: "A: We have no control over our actions." "B: We have some control over our actions." Under A, Harris's argument would read: "We have no control over our actions, therefore we should not use retributive punishment." Wait... we have no control, therefore we should? We can't should anything, if we have no control over our actions. So let's try the other one: "We have control over our actions that isn't 'free will,' therefore we should not use retributive punishment." And here the question inevitably arises: Well, why not?

Coyne and Harris actually seem to think A, which is why I always shake my head when I read their posts about free will. They say things like "A chicken makes decisions; do chickens have free will?" YES! Dammit man, what kind of anthropocentrist are you? Of course a vertebrate has free will, that shouldn't even be up for debate honestly. The proper question is about arthropods--or maybe even bacteria. I think I have made very clear that I believe volition is an evolutionary trait we share with many animals. (The same goes for Dennett and Yudkowsky; they'd be the two modern compatibilists to talk about, someday I hope to be the third.)

Yet, if they really do mean that, we must take them to be denying faculties that we obviously do have. A human being can make decisions based on a predictive model of the world. A human being can respond to arguments and incentives. A human being can be persuaded, cajoled, seduced, pressured, encouraged, rewarded, or coerced. Your model of a human must, to be accurate, include very complex interactions between an enormous array of physical components; and actually, to be remotely tractable, it really needs to include higher-order systems like beliefs, intentions, desires, and emotions.

A rock, on the other hand, cannot do these things. You can push a rock, or melt it, or blow it up; but that's about it. You are limited to very simple physical actions that can be expressed in a very simple physical model---momentum, mass, that sort of thing.

Is the human mind ultimately reducible to such physical properties? We have reason to think so, but the result would be so mind-bogglingly complex that a computer the size of the universe couldn't process all the data in a billion years. The sense in which humans are "reducible" to atoms is the sense in which a computer program is "reducible" to 0s and 1s. (It's almost trivial, frankly; I don't know why anyone doubts it, or why people care so much about it. Yes, we are made of stuff. What else would we be made of?)

Is there some mysterious non-physical substance? Of course not; what would that even mean? But humans are not rocks. We do things rocks don't. Important things--morally important things. And frankly I don't see the problem with calling these special faculties we (and some animals, maybe robots, and not much else) have "free will," but if that has bad connotations for you, by all means, call it something else. I have proposed several alternatives on several occasions: Rational volition, cognition, decision-making processes, behavior algorithms, intelligence... The point is, it makes a difference. You can and should treat an intelligent being different from a mindless physical system. Indeed, rewards and punishments are a very big part of that difference: Intelligent beings respond to incentives and can be operant-conditioned.

Determinism, on the other hand, does not make a difference. We could have no randomness at all, or an awful lot of randomness, and it would not make any difference as far as whether punishment makes sense. (We couldn't have total randomness, because then we couldn't survive at all. This is the "required" part of Yudkowsky's "requiredism".) You don't punish a rock, you don't punish a die roll. Could you punish a robot? Sure, if the robot was configured to learn from punishments (as we are).

There are many legitimate questions to be asked about our justice system, and many important reforms worth making. Psychology and cognitive science have a lot to say about that. But it just doesn't have a damn thing to do with free will.