codex Slate Star Codex

NꙮW WITH MꙮRE MULTIꙮCULAR ꙮ

To Understand Polarization, Understand Conservativism’s Failures

Earlier today I talked about one reason for increased polarization on the Democratic side. Now I want to match it with one reason the Republicans have even more of a problem.

Ask anyone what Republicans want, and they’ll say things like “smaller government”, “fewer regulations”, and “less welfare state”.

Meanwhile, here are some graphs showing how they’re doing (disclaimer: graphs like this are very dangerous, and I can only plead that I’ve seen numbers like these from enough sources that I think they have some contact with reality):

Source: https://fivethirtyeight.blogs.nytimes.com/2013/01/16/what-is-driving-growth-in-government-spending/

Source: https://regulatorystudies.columbian.gwu.edu/reg-stats

Source: http://www.motherjones.com/kevin-drum/2016/02/heres-why-bernie-sanders-doesnt-say-much-about-welfare-reform/

Apparently not so good.

This is true even though this is a historic apex of Republican power. They control the House, the Senate, the Presidency, 66% of state governorships, 68% of relevant state legislatures, and are kind of tied-ish for control of the Supreme Court. They’ve been two of the last four Presidents, and controlled Congress more often than not during that period.

This is really strange. Whatever they wanted, they should have been able to get. Who’s going to stop them? Democrats? Don’t make me laugh.

But in fact, we mostly kept getting bigger government, more regulations, and a bigger welfare state.

My guess is this is a larger-scale version of what I talked about in Considerations On Cost Disease. Various secular trends make everything more expensive and worse, which means government has to spend more money and regulation to get the same level of services, which means government gets bigger. There’s no easy way to stop this except to understand cost disease (which people don’t) or to drastically cut the level of services and admit it will keep getting worse (which politicians are scared of doing on their watch). This is not really the Republicans’ fault.

But Republican voters don’t know that.

All they see is candidates running for office on a platform of small government and less regulation. Then they win, they’ve got a huge majority and a great mandate, and at the end of their term government is as big as always and there are more regulations than ever.

And if maybe you’re not that sophisticated about these kinds of things, you think – these guys betrayed me. They’re Republicans in name only. They were corrupted by Washington. The liberal media finally got them. They’re weak and they caved as soon as the Democrats called them mean names. What we need are some real Republican candidates, ones who are actually willing to stand up to the establishment.

Then you elect the Real Republican Candidates, the Tea Party or whoever, and the same thing happens. Because we’re talking about secular trends and not about anything that Congress can easily affect.

So then the voters think they’re frauds too, and they get defeated in the primaries by other people who are even more Tea Party than they are, people who can say oh yeah, those Tea Party people were fake, but we have the necessary commitment to go to Washington and not cave in immediately.

This will never work. But the superficial logic of “Republicans are powerful enough to get whatever they want, we don’t have small government, therefore the current crop of so-called Republicans didn’t really want small government enough” is convincing. You end up with a signaling spiral where everyone’s in an arms race to show that they’re not actually the craven compromisers that people will inevitably assume them to be. That means hyperpartisanship and refusal to compromise on anything.

I’m talking about this as a Republican problem, but I think it’s a general issue whenever people have unrealistic expectations, ie always. The more our hopes diverge from the possible, the more we’ll reject all existing governments in favor of stronger and stronger forms of extremism.

Posted in Uncategorized | Tagged | 604 Comments

Against Murderism

[Content warning: discussion of racism. Comments are turned off due to bad experience with the comments on this kind of material.]

I.

A set of questions, hopefully confusing:

Alice is a white stay-at-home mother who is moving to a new neighborhood. One of the neighborhoods in her city is mostly Middle Eastern immigrants; Alice has trouble understanding their accents, and when they socialize they talk about things like which kinds of hijab are in fashion right now. The other neighborhood is mostly white, and a lot of them are New Reformed Eastern Evangelical Episcopalian like Alice, and everyone on the block is obsessed with putting up really twee overdone Christmas decorations just like she is. She decides to move to the white neighborhood, which she thinks is a better cultural fit. Is Alice racist?

Bob is the mayor of Exampleburg, whose bus system has been losing a lot of money lately and will have to scale back its routes. He decides that the bus system should cut its least-used route. This turns out to be a bus route in a mostly-black neighborhood, which has only one-tenth the ridership of the other routes but costs just as much. Other bus routes, most of which go through equally poor mostly-white neighborhoods, are not affected. Is Bob racist?

Carol is a gay libertarian who is a two-issue voter: free markets and gay rights. She notices that immigrants from certain countries seem to be more socialist and more anti-gay than the average American native. She worries that they will become citizens and vote for socialist anti-gay policies. In order to prevent this, she supports a ban on immigration from Africa, Latin America, and the Middle East. Is Carol racist?

Dan is a progressive member of the ACLU and NAACP who has voted straight Democrat the last five elections. He is studying psychology, and encounters The Bell Curve and its theory that some of the difference in cognitive skills between races is genetic. After looking up various arguments, counterarguments, and the position of experts in the field, he decides that this is probably true. He avoids talking about this because he expects other people would misinterpret it and use it as a justification for racism; he thinks this would be completely unjustified since a difference of a few IQ points has no effect on anyone’s basic humanity. He remains active in the ACLU, the NAACP, and various anti-racist efforts in his community. Is Dan racist?

Eric is a restauranteur who is motivated entirely by profit. He moves to a very racist majority-white area where white people refuse to dine with black people. Since he wants to attract as many customers as possible, he sets up a NO BLACKS ALLOWED sign in front of his restaurant. Is Eric racist?

Fiona is an honest-to-goodness white separatist. She believes that racial groups are the natural unit of community, and that they would all be happiest set apart from each other. She doesn’t believe that any race is better than any other, just that they would all be happier if they were separate and able to do their own thing. She supports a partition plan that gives whites the US Midwest, Latinos the Southwest, and blacks the Southeast, leaving the Northeast and Northwest as multiracial enclaves for people who like that kind of thing. She would not use genocide to eliminate other races in these areas, but hopes that once the partition is set up races would migrate of their own accord. She works together with black separatist groups, believing that they share a common vision, and she hopes their countries will remain allies once they are separate. Is Fiona racist?

II.

As usual, the answer is that “racism” is a confusing word that serves as a mishmash of unlike concepts. Here are some of the definitions people use for racism:

1. Definition By Motives: An irrational feeling of hatred toward some race that causes someone to want to hurt or discriminate against them.

2. Definition By Belief: A belief that some race has negative qualities or is inferior, especially if this is innate/genetic.

3. Definition By Consequences: Anything whose consequence is harm to minorities or promotion of white supremacy, regardless of whether or not this is intentional.

Some thoughts:

Definition By Consequences Doesn’t Match Real-World Usage

I know that Definition By Consequences is the really sophisticated one, the ones that scholars in the area are most likely to unite around. But I also think it’s uniquely bad at capturing the way anyone uses the word “racism” in real life. Let me give four examples.

First, by this definition, racism can never cause anything. People like to ask questions like “Did racism contribute to electing Donald Trump?” Under this definition, the question makes no sense. It’s barely even grammatical. “Did things whose consequence is harm minorities whether or not such harm is intentional contribute to the election of Donald Trump?” Huh? If racism is just a description of what consequences something has, then it can’t be used as an causal explanation.

Second, by this definition, many racist things would be good. Suppose some tyrant wants to kill the ten million richest white people, then redistribute their things to black people. This would certainly challenge white supremacy and help minorities. So by this definition, resisting this tyrant would be racist. But obviously this tyrant is evil and resisting him is the right thing to do. So under this definition, good policies which deserve our support can nevertheless be racist. “This policy is racist” can no longer be a strong argument against a policy, even when it’s true.

Third, by this definition, it doesn’t make a lot of sense to say a particular person is racist. Racism is a property of actions, not of humans. While there are no doubt some broad patterns in people, the question “Is Bob racist?” sounds very odd in this framework, sort of like “Does Bob cause poverty?” No doubt Bob has done a few things which either help or hurt economic equality in some small way. And it’s possible that Bob is one of the rare people who organizes his life around crusading against poverty, or around crusading against attempts to end poverty. But overall the question will get you looked at funny. Meanwhile, questions like “Is Barack Obama racist?” should lead to a discussion of Obama’s policies and which races were helped or hurt by them; issues like Obama’s own race and his personal feelings shouldn’t come up at all.

Fourth, by this definition, it becomes impossible to assess the racism of an action without knowing all its consequences. Suppose the KKK holds a march through some black neighborhood to terrorize the residents. But in fact the counterprotesters outnumber the marchers ten to one, and people are actually reassured that the community supports them. The march is well-covered on various news organizations, and outrages people around the nation, who donate a lot of money to anti-racist organizations and push for stronger laws against the KKK. Plausibly, the net consequences of the march were (unintentionally) very good for black people and damaging to white supremacy. Therefore, by the Sophisticated Definition, the KKK marching the neighborhood to terrorize black residents was not racist. In fact, for the KKK not to march in this situation would be racist!

So Definition By Consequences implies that racism can never be pointed to as a cause of anything, that racist policies can often be good, that nobody “is a racist” or “isn’t a racist”, and that sometimes the KKK trying to terrorize black people is less racist than them not trying to do this. Not only have I never heard anyone try to grapple with these implications, I see no sign anyone has ever thought of them. And now that I’ve brought them up, I don’t think anyone will accept them as true, or even worry about the discrepancy.

I think this is probably because it’s a motte-and-bailey, more something that gets trotted out to win arguments than anything people actually use in real life.

Definition By Belief Is A Mess

Is it racist to believe that Mexicans are poorer than white people? After all, being poor is generally considered bad, so you’re attributing a bad quality to a minority group. What if you add “Mexicans are only poor because of being oppressed and discriminated against?”

Is it racist to believe that Mexicans are more criminal than white people? What if you add “Mexicans are only criminal because their culture was shaped by the experience of oppressive Spanish colonization, which left deep scars on their national psyche”?

Is it racist to believe that Muslims commit more terrorism than white people? What if you’ve done a lot of calculations of per capita terrorist attacks and you can quote exact numbers that prove your point?

Is it more or less racist if then you add “…but this is because Islam is a violent religion that encourages murder, and has nothing to do with the genetics of Middle Eastern people”?

Is it racist to believe that Pygmies are shorter than white people?

But None Of That Really Matters, Because In Real Life, Definition By Motive Usually Trumps Definition By Belief

After the London attacks, I heard someone ask “Do you have to be a racist to want to restrict immigration from Muslim countries? Or can you just be really worried about the terrorism risk?”

A lot of people responded. Some of them said no, it was perfectly reasonable to be worried about terrorism. Other people said that concern about terrorism was just a smokescreen, that people said they were just concerned about terrorism, but actually that was just a way to cover up their racism.

Think about how confusing this is. It’s positing two different things. First, a belief that Muslims are often terrorists and so we should crack down on them. And second, racism. These things are considered opposing explanations, such that if we believe the first one, we can dismiss the second – or, if we admit the second, that proves the first was claimed dishonestly. Under Definition By Belief, it’s really weird.

(compare: “I believe Jesus Christ is the Son of God and those who believe in Him gain eternal life.” “No, you’re just using that as a smokescreen to cover up that you’re Christian!”)

The only way I can make sense of this argument is to think of it as Definition By Motive trumping Definition By Belief. The first person is stating a belief that Muslims are more likely to be terrorists. The second person is questioning whether their motivation for restricting immigration is really this belief (in which case it would be ok) or if they’re motivated by an irrational hatred of minorities (in which case it would be racism).

Definition By Motive can even trump Definition By Belief when we’re talking about innate/genetic difference. Consider Charles Murray saying that he believes black people are genetically less intelligent than white people. Some of Murray’s critics object that this should be suppressed, even if true, because it could be used to justify racism.

Under Definition By Belief, this makes no sense. Imagine Murray was a geologist, pointing out that Antarctica contained mostly sedimentary rock. His critics object “This should be suppressed, even if true, because it could be used to justify believing that Antarctica contains mostly sedimentary rock!” Huh?

It makes more sense if we think of it as being about Definition By Motive. Then the critics are saying that if we find that minority groups are genetically worse in some way (ie racist beliefs are true), we should suppress that lest it be used to justify people’s irrational feelings of hatred for members of other races (ie justify racist motives).

Definition By Motive Makes Sense Of All Of The Above Examples And Basically Matches Most Real-World Usage

Definition By Motive fits the first example. When we ask “Was racism responsible for Trump’s election?” we mean “Did people elect Trump because they irrationally hated minorities and wanted to discriminate against them?”

It fits the second example. When we say that it wouldn’t be racist to resist a tyrant who wants to kill whites, we mean that such resistance is a good policy, which would be pursued for reasons other than just irrationally hating minorities and wanting to discriminate against them.

It fits the third example. When we claim a specific person (Bob, Barack Obama) is racist, we mean that they irrationally hate minorities and want to discriminate against them.

It fits the fourth example. When we say that the KKK marching through a black neighborhood to terrorize people is racist regardless of its consequences, we mean that it’s motivated by an irrational hatred of minorities and desire to discriminate against them.

It fits the fifth example. When we ask whether it’s racist to believe Mexicans are poorer than whites, we’re asking whether someone would only say that because they irrationally hate Mexicans and want to discriminate against them. But most of the time people making that claim are trying to point out inequalities and help Mexicans. So it isn’t racist.

It fits the sixth example. Somebody who believes that Mexicans are more criminal than white people might just be collecting crime stats, but we’re suspicious that they might use this to justify an irrational hatred toward Mexicans and desire to discriminate against them. So it’s potentially racist, regardless of whether you attribute it to genetics or culture.

It fits the seventh example. It’s probably not racist to believe that Muslims commit more terrorism than white people, since this seems to be a true or at least plausible claim, but if people talk about it too much it’s worrying that maybe they’re trying to justify their irrational hatred of Muslims and desire to discriminate against them.

It fits the eighth example. It’s probably not racist to believe that Pygmies are shorter than white people, because it’s obviously true and you would believe it whether you had an irrational hatred of Pygmies or not. Also, no one cares how tall anybody is.

It fits the ninth example. When people ask whether immigration restricts are really due to fear of terrorism vs. racism, they’re asking whether people who claim to be concerned with terrorism actually just irrationally hate minorities and want to discriminate against them.

And it fits the tenth example. When people say that Charles Murray’s claims about genetics might be used to “justify racism”, they mean that if you irrationally hate minorities and want to discriminate against them, you could use his claims about genetics as a justification for why your position makes sense.

Overall We Probably Use A Combination Of All Of These, Weighted In Favor Of Definition By Motives

I designed the discussion questions to be situations where Definition By Motive clearly didn’t apply, but one or both of the other definitions clearly did. I imagine some people stuck to their guns, went Definition By Motive all the way through, and said none of the people in the vignettes were racist. I imagine other people used one of the other two definitions, or a different definition of their own, and were able to navigate all of the objections and counterexamples down here in Part II successfully. But I think most people found a couple of inconsistencies, aren’t really sure what to do with them, and are just sort of echoing the Supreme Court’s view of pornography: “I’ll know it when I see it.”

This is natural. I’m not trying to say that Definition By Motives is the one “real” definition. All of our word usage is a mess; we hardly ever use anything simply or consistently, let alone a complicated word like “racism”. In reality we go back and forth among all of these, proving that something is racist using one definition, then applying the consequences of another definition, switching from very strict to very loose based on whether or not it’s something we like. All of this is totally normal.

But in this case it’s kind of likely to lead to disaster.

III.

A digression, from an alternative universe.

“Murderism” is the ideology that murdering people is good and letting them live is bad. It’s practically omnipresent: 14,000 people are murdered in the US each year. That’s a lot of murderists, and a testament to the degree to which our schools teach murderist values.

But not all murderism is that obvious. For years, people have been pushing “soft-on-crime” policies that will defund the police and reduce the length of jail sentences – inevitably increasing the murder rate. Advocates of these policies might think that just because they’re not gangsters with knives, they must not be murderists. But anybody who supports murder, whether knife-wielding gangster or policy analyst – is murderist and responsible for the effects of their murderism.

Our two major parties have many differences – but both are united in their support for murderism. Republicans push murderist policies like the invasion of Iraq, which caused the murder of thousands of Iraqis. Democrats claim to be better, but they support openly murderist ideas like euthanasia, promoting the killing of our oldest and most vulnerable citizens. There’s no party in Washington that’s willing to take a good look at itself and challenge the murderist ideals that our political system is built on.

Murderism won’t stop until people understand that it’s not okay to be murderist. So next time you hear people opposing police militarization, or speaking out in favor of euthanasia – tell them that that’s murderism and it’s not okay.

…okay, done. Back in our own universe, we recognize that “murderism” is silly: it confuses cause and effect.

Murder is usually an effect of a strategy pursued for other reasons. The drug dealer who wants to keep rivals off his turf, the soldier who wants to win a war, the gangster who wants to get rid of inconvenient witnesses. If you want to stretch it, add the neocon who wants to “liberate” foreign countries, the cancer patient who wants to “die with dignity”, or the activist who wants to keep people out of jail.

But except in maybe the most deranged serial killers, it’s never pursued because of an inherent preference for murder. Most murderers would probably prefer not to have to kill. If the drug dealer could protect his business equally well by politely requesting people stay off his territory, that would be much easier. If the soldier could win his war without bloodshed, so much the better for everybody. Murder is an effect of other goals – sometimes base, sometimes noble – and the invocation of “murderism” only serves to hide these goals and conflate different actions into a single meaningless category.

Talking about murderism isn’t just uninformative, it’s actively confusing. If you believed that gangsters killed their rivals because of murderism, then there’s no point in examining how poverty interacts with gang membership, or whether the breakdown of law forces people to form gangs to defend themselves. The problem is just that gangsters have murderist values. It should be solved by censoring the works of philosopher David Benatar, who writes about how being alive is bad and it’s morally better not to exist at all. Or by banning high school Goths, whose pro-death aesthetic makes murderism seem cool to teens and causes them to harbor murderist thoughts as adults.

Talk about murderism is obviously confused. But it’s the same confusion between the Definition By Consequences versus the Definition By Motive that we saw was a hallmark of racism.

IV.

Belief in murderism creates a hostile and ineffective society whose weird beliefs can only be countered by accepting that murder is rarely a terminal goal, but a usually result of strategies pursued for other reasons. We accept that having a terminal goal of killing people seems so awful, inhuman, and incongruous with the sort of decent humans we all know – that it’s a very strange explanation to even consider when other, better ones are available. We can apply the same analysis to racism. The discussion questions in Part I already started the process, but we can go further.

I’m not just making the argument “lots of things aren’t really racist”. I can’t do much about how you choose to define words, plus it’s doomed to fail anyway. Imagine having to convince someone that a guy who committed homicide “isn’t really murderist”. Doesn’t sound like the most winnable fight.

And if you only break down non-racist things into non-racist motives, what reward shall you have? Do not even the scribes and the Pharisees do the same? I say unto you, if you want to be righteous, look for the non-racist motives in actually racist things.

What does that mean?

Consider some business, let’s say a daycare center, that we know discriminates against black job-seekers. If we ask them why, they say “Because black people are criminal”. This sounds like just about the most typical and obvious example of racism possible.

But there’s actually a lot of really good scholarship on this exact situation, and it helps provide a different perspective. It starts like this – a while ago, criminal justice reformers realized that mass incarceration was hurting minorities’ ability to get jobs. 4% of white men will spend time in prison, compared to more like 16% of Hispanic men and 28% of black men. Many employers demanded to know whether a potential applicant had a criminal history, then refused to consider them if they did. So (thought the reformers) it should be possible to help minorities have equal opportunities by banning employers from asking about past criminal history.

The actual effect was the opposite – the ban “decreased probability of being employed by 5.1% for young, low-skilled black men, and 2.9% for young, low-skilled Hispanic men.”

In retrospect, this makes sense. Daycare companies really want to avoid hiring formerly-imprisoned criminals to take care of the kids. If they can ask whether a certain employee is criminal, this solves their problem. If not, they’re left to guess. And if they’ve got two otherwise equally qualified employees, and one is black and the other’s white, and they know that 28% of black men have been in prison compared to 4% of white men, they’ll shrug and choose the white guy.

Is this racist? Is this “statistical discrimination”? Describe it with whatever word you want. The point is that they have understandable motives (don’t hire criminals to take care of the kids), accurate beliefs, and in their shoes you might do the same. More important, once you give them the tools they need to solve their problems without racial discrimination – you let them see applicants’ criminal histories – they have no further desire to discriminate and your problem is solved.

If you tried to solve this by sending these people to sensitivity training, you would fail. IF you tried to solve this by firing these people, then the people who replaced them would have the same incentives, and you would fail again. If you try to solve it by realizing that racial animus has no role at all in this scenario, and daycare owners just want to do what’s best for their kids, then you can provide them with the tools they need to do that, and solve the racial discrimination at the same time.

Okay, fine. Harder example. Let’s take, uh, some guy who’s always ranting about how the Jews secretly control the world. They have underground tunnels where they have their secret Zionist meetings and talk about how they’re going to stick it to the Christians. Every major war and economic downturn has been caused by this. Are we allowed to treat this guy‘s racism as being a conceptual primitive that doesn’t need further breaking down?

I actually knew a guy like this. He was a schizophrenic patient in the mental hospital where I work. Overall I found it a nice break from the tedium of CIA-conspiracy folk, alien-conspiracy folk, and white-people-conspiracy folk (remember, this is Detroit).

Am I saying everyone like this is schizophrenic? Not diagnosably, no. But I notice that there are a lot of not-diagnosably-schizophrenic people who believe in the Illuminati, the New World Order, the Freemasons, and – yes – lizardmen. Is it really so outlandish to say that the same faulty reasoning that concludes that Freemasons run the world could conclude that Jews run the world, and for the same reasons? Does it really make sense to just blow one off as paranoid conspiracy-mongering, and the other as originating from a completely different process called “anti-Semitism” or “racism”? Remember, “healthy” people with paranoid and conspiratorial beliefs have the same kind of fronto-striatal prediction error signal that schizophrenics do, only less so, suggesting that their odd ideas probably come from the same kind of disturbed reasoning process.

“Are you saying that anti-Semitism literally plays no role in their theory about the Elders of Zion”? Again, call it what you want. I’m saying that by totally ignoring the anti-Semitic aspect, I was able to successfully treat this guy with Seroquel, whereas if you tried to read him Elie Wiesel books, he’d still be in that psych ward today.

Fine. Schizotypal conspiracy-mongerers are a noncentral example anyway. What about, I don’t know, rural Republicans in South Carolina who wave the Confederate flag all the time and think blacks and immigrants are ruining the country.

Here I would point out that this is pretty much the demographic that elected Nikki Haley (birth name, Nimrata Randhawa; daughter of two Punjabi immigrants) as governor, and that supports her so fervently that she remains one of the most popular politicians in the country. Also the demographic that loved Ben Carson, making him the only candidate to briefly displace Trump for first place in the 2016 Republican primary polls. One plausible explanation is that the South Carolinians don’t like blacks and immigrants because they view them as having foreign values – specifically, Blue Tribe values (it may be relevant here that 90%+ of blacks usually vote Democrat). If someone like Nikki Haley or Ben Carson proves that they share Red beliefs, they become part of the tribe and will be fiercely defended. Maybe this is more like the daycare situation than it looks – people using race as a proxy for something they care about, until they get direct information.

To be clear – I am not saying that racism doesn’t exist, I’m not saying that we should ignore racism, I’m not saying that minorities should never be able to complain about racism. I’m saying that it’s very dangerous to treat “racism” as a causal explanation, that it might not tell you anything useful about the world, and that’s a crappy lever to use if you want to change behavior.

And I’m not saying that it’s not useful to think of some of these things as places where there’s an opportunity for racial change. If a daycare owner is really interested in redressing racial inequality, they can hire minorities even if it’s against their incentives and self-interest (although it’s unclear why the owner should prefer that opportunity to other opportunities, like donating some of their profits to the NAACP.)

And I’m not saying that there will never be a case that’s impossible to break down into non-racist motives. Heck, I’m not even saying there aren’t some honest-to-goodness murderists out there. But I am saying we should at least try. Not because it’s necessarily costless. Not because there isn’t a risk of false negatives.

We should try because it’s the only alternative to having another civil war.

V.

Arnold Kling likes to talk about how political groups are divided by different “languages”, different schemata for understanding the world that make it difficult to talk across political divides.

Jonathan Haidt accepts the premise but challenges the symmetry; his experiments ask liberals and conservatives to fill out questionnaires about their values, then to predict how someone from the opposite tribe would fill out the questionnaire. He finds that conservatives are able to predict liberals’ answers just fine and seem to have a pretty good understanding of their worldviews, but that liberals have no idea how conservatives think or what they value.

James Scott, as channeled by Lou Keep, draws the asymmetry a little differently. He says that the process of development, especially state-building and the switch from traditional to market economies, creates a pressure for “legible” language that renders entire classes of problems very difficult to talk about. This creates an asymmetry between an elite plugged into the global market structure whose concerns make perfect sense (“If we do this, GDP will go up 3% and we can build more roads!”) and the masses left behind whose concerns seem pointless and vague (“I feel like something important disappeared when we turned everything into a commodity”). Keep then proposes a very loose mapping onto cosmopolitan neoliberal Clintonites versus undereducated “I’m angry about losing my traditional culture” Trumpists.

There are a bunch more frameworks like this, but they all share the common warning that cross-cultural communication is really hard, and so a lot of the concerns of people who aren’t like us will probably sound like nonsense. And most of them say that our demographic – well-educated people proud of our commitment to logic and reason – are at especially high risk of just dismissing everyone else as too dumb to matter. The solution is the same as it’s always been: hard work, renewed commitment to liberal values, and a hefty dose of the Principle of Charity.

Racism-as-murderism is the opposite. It’s a powerful tool of dehumanization. It’s not that other people have a different culture than you. It’s not that other people have different values than you. It’s not that other people have reasoned their way to different conclusions from you. And it’s not even that other people are honestly misinformed or ignorant, in a way that implies you might ever be honestly misinformed or ignorant about something. It’s that people who disagree with you are motivated by pure hatred, by an irrational mind-virus that causes them to reject every normal human value in favor of just wanting to hurt people who look different from them.

This frees you from any obligation to do the hard work of trying to understand other people, or the hard work of changing minds, or the hard work of questioning your own beliefs, or the hard work of compromise, or even the hard work of remembering that at the end of the day your enemies are still your countrymen. It frees you from any hard work at all. You are right about everything, your enemies are inhuman monsters who desire only hatred and death, and the only “work” you have to do is complain on Twitter about how racist everyone else is.

And I guess it sounds like I’m upset that we’re not very good at solving difficult cross-cultural communication problems which require deep and genuine effort to understand the other person’s subtly different value system. I’m not upset that we can’t solve those. Those are hard. I’m upset because we’re not even at the point where someone can say “I’m worried about terrorism,” without being forced to go through an interminable and ultimately-impossible process of proving to a random assortment of trolls and gatekeepers that they actually worry about terrorism and it’s not just all a ruse to cover up that they secretly hate everyone with brown skin. I’m saying that when an area of the country suffers an epidemic of suicides and overdoses, increasing mortality, increasing unemployment, social decay, and general hopelessness, and then they say they’re angry, we counter with “Are you really angry? Is ‘angry’ just a code word for ‘racist’?” I’m saying we’re being challenged with a moonshot-level problem, and instead we’re slapping our face with our own hand and saying “STOP HITTING YOURSELF!”

People talk about “liberalism” as if it’s just another word for capitalism, or libertarianism, or vague center-left-Democratic Clintonism. Liberalism is none of these things. Liberalism is a technology for preventing civil war. It was forged in the fires of Hell – the horrors of the endless seventeenth century religious wars. For a hundred years, Europe tore itself apart in some of the most brutal ways imaginable – until finally, from the burning wreckage, we drew forth this amazing piece of alien machinery. A machine that, when tuned just right, let people live together peacefully without doing the “kill people for being Protestant” thing. Popular historical strategies for dealing with differences have included: brutally enforced conformity, brutally efficient genocide, and making sure to keep the alien machine tuned really really carefully.

And when I see someone try to smash this machinery with a sledgehammer, it’s usually followed by an appeal to “but racists!”

You say we must protect freedom of speech. But would you protect the free speech of racists?

You say people shouldn’t get fired for personal opinions that don’t affect their work. But would you let racists keep their jobs?

You say we try to solve disagreements respectfully through rational debate. But would you try to rationally debate racists?

You say people should be allowed to follow their religion without interference. But what if religion is just a cover for racism?

You say we need to understand that people we disagree with can sometimes have some good points. Are you saying we should try to learn things from racists?

You say there’s a taboo on solving political disagreements by punching people. Are you saying that we can’t punch racists?

The argument goes: liberalism assumes good faith and shared values. It assumes that, at the end of the day, whether you’re Catholic or Protestant, you can still be a basically good person. You can compartmentalize a few special beliefs relating to the Pope, and your remaining differences can be dissolved by the universal solvent of Reason. After everyone does this, you can invoke the wisdom of crowds via a popular election, and even if you don’t like the results you can at least understand where the other side is coming from. Some people prefer liberty to safety, other people prefer safety to liberty, but if the voters choose the wrong one then at least they’ve erred in an understandable way by preferring one real value to another.

But if there’s some group out there who aren’t connected to normal human values at all, some group that’s deliberately rejected reason; if they’re willing to throw liberty and safety under the bus in pursuit of some kind of dark irrational hatred which is their only terminal goal – then the whole project falls apart. Dialogue based on mutual trust may be all nice and well when it’s supposed to help us choose the optimal balance between liberty and safety, but if you give a platform to the people whose only value is hatred, then they’re just screwing over everybody.

A few days ago, Noah Smith posted on Twitter about hearing some people say racist things. The comments went like this:

Ah well. They said a racist thing. Guess we’ve got to kill them.

And I agree with this chain of logic. Using violence to enforce conformity to social norms has always been the historical response. We invented liberalism to try to avoid having to do that, but you can’t liberalism with people who refuse reason and are motivated by hatred. If you give the franchise to green pointy-fanged monsters, they’re just going to vote for the “Barbecue And Eat All Humans” party. If such people existed and made up a substantial portion of the population, liberalism becomes impossible, and we should go back to just using violence to enforce our will on the people who disagree with us. Assuming they don’t cooperate with our strategy of violently suppressing them, that means civil war.

I don’t want civil war. I want this country to survive long enough to be killed by something awesome, like AI or some kind of genetically engineered superplague. Right now I think going out in a neat way, being killed by a product of our own genius and intellectual progress – rather than a product of our pettiness and mutual hatreds – is the best we can hope for. And I think this is attainable! I think that we, as a nation and as a species, can make it happen.

But it starts with rejecting the “murderism” framework. Rejecting the choice to attribute whatever we disagree with to murderism, even if it is murderist, and instead trying to trace it back to root causes that make sense that and humanize the people involved. Working to find the reasons liberalism is possible, rather than the reasons it isn’t. Unless we can do that, semantic confusion and our political polarization are going to build off each other in a vicious cycle into who knows where.

Posted in Uncategorized | Tagged , | Comments Off on Against Murderism

OT78: Oprah Thread

This is the bi-weekly visible open thread. Post about anything you want, ask random questions, whatever. You can also talk at the SSC subreddit, the SSC Discord server, or the Cafe Chesscourt forum.

Posted in Uncategorized | Tagged | 1,288 Comments

Links 6/17: Silinks Is Golden

Did you know: JRR Tolkien’s great granddaughter, Ruth Tolkien, is the only blind person in the UK to be a competitive fencer. She is currently ranked the #186th best fencer in the country.

Alpha – an ambiguously-AI-automated alliterative account about abiogenesis. As an annotator already announced, “absolutely an amazing achievement”.

In response to my Silicon Valley reality check, Noah Smith looks for good critiques of Silicon Valley.

Sarah C is interested in a potential sepsis cure and pleads for you to consider helping a hospital fund a study. Outside my area of expertise, except to say that sepsis is really bad and curing it would be pretty great, plus I trust Sarah.

Neerav Kingsland: Ignoring Educational Productivity Is Immoral: “studies consistently find urban charter schools outperform traditional public schools while spending around 20% less per-pupil…what, as a society, could we do with this 20% extra funding that urban charter schools could save us?” I would love to see a debate between Neerav Kingsland and Freddie deBoer – I can host or otherwise try to make it happen if they agree.

Related – Wanted: A Charter High School That Starts Class At A Reasonable Hour

New front on the battle to prove that zero-calorie sodas must be bad in some way: does CO2 in carbonated beverages induce ghrelin release and increase food consumption?

According to Muslim legend, the Jews will try to hide behind trees to escape the Last Judgment, but the trees will shout “There is a Jew behind me, come and kill him!” – except for the Gharqad tree, which is apparently pro-Jew. This has led to all sorts of amazing conspiracy theories, like that Israel is planting a bunch of Gharqad trees to ensure hiding places for its citizens. And here’s a book about Hamas terrorists with an interesting passage on the Gharqad tree legend.

A paper claims that housing restrictions have “lowered aggregate US growth by more than 50% from 1964 to 2009″.

Explaining The Gender Gap In Crime: The Role Of Heart Rate. “A low resting heart rate is widely regarded as the best-replicated biological correlate of antisocial and criminal behavior” Apparently it might have to do with something general level of autonomic arousal being lower in (to be blunt about it) emotionless sociopaths?

The best Internet responses to Trump touching a glowing orb in Saudi Arabia.

I don’t have a source for this, but it looks right, and it’s a really cool way of presenting the data:

The mathematics of Girih tiles, a beautiful form of Islamic art which applied principles of Penrose tilings five hundred years before it was discovered in the West.

Vox tries its hand at an explainer about the Sam Harris / Charles Murray interview. Some criticism from Gene Expression, The Misrepresentation Of Genetic Science In The Vox Piece On Race And IQ. From Elan, The Cherry-Picked Science In Vox’s Charles Murray Article. From Sam Harris, an accusation that the article just blatantly lies about the contents of the publicly available podcast (one of the authors later apologizes for this, but Vox hasn’t changed the article). From Professor Emeritus Richard Haier, who called it a “junk science piece” and tried to write a counterpiece for Vox (they refused to publish it, but it’s now up on Quillette). And even from other Vox reporters who thought it was journalistically shoddy. As for me, I think the article was as good as it could be under the circumstances – while it does get some things wrong and is personally unfair to Murray, from a scientific point of view I’m just really glad that the piece admits that IQ is real, meaningful, and mostly hereditary. This was the main flashpoint of the original debate twenty-five years ago, it’s more important than the stuff on the achievement gap, and the piece gets it entirely right. I think this sort of shift from debating the very existence of intelligence to debating the details is important, very productive, and worth praising even when the details are kind of dubious. This should be read in the context of similar recent articles like NYMag’s Yes, There Is A Genetic Component To Intelligence and Nature’s Intelligence Research Should Not Be Held Back By Its Past.

AskHistorians: Did Roman legionnaires get PTSD? “For the Romans, people experiencing intrusive memories were said to be haunted by ghosts…those haunted by ghosts are constantly depicted showing many symptoms which would be familiar to the modern PTSD sufferer.”

Somehow I went through medical school without ever learning that going in hot tubs while pregnant can be a risk factor for birth defects.

Like a food blog, except it’s RPG rations for dwarves, elves, orcs, et cetera.

Highly educated people are more likely to get brain tumors. The article mentions the boring hypothesis that they just have better access to medical care (but then how come most other cancers are higher in the uneducated?), but also proposes the much more interesting hypothesis that “having more brain cells or greater brain activity somehow increases a person’s risk”.

Future of Humanity Institute in Oxford is looking for a research assistant for Toby Ord (founder of effective altruism, currently working on a book on existential risks). If you’re interested, check out the job details and contact info.

According to the Guinness Book Of World Records, the most fraudulent election in history was the 1927 Liberian presidential contest, in which incumbent Charles King received 234,000 votes despite there being only 15,000 registered voters.

The “moderate drinking increases lifespan” vs. “that’s obviously just a confounder based on only healthy people drinking” wars continue, with the latest volley being that fruit flies and chickens exposed to alcohol vapor live longer. I was previously on the “obviously just a confounder” side of the debate, but the animal studies sound pretty convincing.

Rudyard Kipling, H.G. Wells, Arthur Conan Doyle, P.G. Wodehouse, G.K. Chesterton, and A. A. Milne all played on the same amateur cricket team. The team called themselves the Allahakbarries “[under] the mistaken belief that Allahu akbar meant ‘Heaven help us’ in Arabic.”

Heredity watch: Elon Musk’s maternal grandparents were well-known pilots and explorers, and Musk’s mother spent part of her childhood on various family expeditions in search of a lost city in the Kalahari Desert.

The best new blog I’ve come across recently is Sam[]zdat, which among other things has been reviewing various great books. Their Seeing Like A State review is admittedly better than mine, but I most appreciated The Meridian Of Her Greatness, based on a review of Karl Polanyi’s The Great Transformation. Go for the really incisive look at important ideas and social trends, stay for the writing style.

What lesson should we draw about Democrats’ prospects from the Republicans’ 7 point win in the Montana special election? (point, counterpoint).

The Less Wrong Wiki hosts a List of Rationalist Podcasts.

Behavioral Individuality In Clonal Fish Arises Despite Near-Identical Rearing Conditions. Worth interpreting in the context of my post Non-Shared Environment Doesn’t Just Mean Schools And Peers – even fish raised in exactly the same environments will show “non-shared environment” effects, probably because of something like embryogenetic randomness.

An easier way to read the Less Wrong Sequences online at readthesequences.com.

Mark Zuckerberg Calls For Universal Basic Income In His Harvard Commencement Speech. Sure, Silicon Valley people call for lots of things, but this seems especially important insofar as Zuckerberg seems like he’s positioning himself to run for President at some point.

An analysis showing Donald Trump’s speech patterns getting less fluent and more bizarre over the past few years – might he be suffering from mild age-related cognitive impairment? Also, given that this can be pretty subtle (cue joke about Trump) and affect emotional stability in complicated ways, should we be more worried about electing seventy-plus year old people to the presidency?

A sobering statistic on the difficulty of dialing back mass incarceration (incompletely sourced, but seems to check out): “if America only jailed murderers and rapists, it would still have more prisoners per capita than Western Europe”.

Measures Of Dogs’ Inhibitory Abilities Do Not Correlate Across Tasks. A lot of good cognition studies are being done on dogs these days; this one suggests that we don’t yet have a good general concept of “inhibition” that we can use to say that some animals (people?) have better self-control in general than others.

Big systematic review and meta-analysis: what actually helps lower-income students succeed? Read the paper or the Freddie deBoer blog post, which summarizes the results as “human beings”.

A more recent study on the evolutionary history of Ashkenazi genetic diseases (paper, popular article) claims that Tay-Sachs is probably the result of evolutionary selection, but that others (eg Gaucher, torsion dystonia, Fanconi anaemia, etc) aren’t, which would be a partial blow for Cochran et al’s selection theory. My guess is they just don’t have enough power to detect the effects – if this was all random drift, it would be vanishingly unlikely that so many of these diseases end up in the same pathways (eg lysosomal storage). [EDIT: I may be misunderstanding this; it may just mean there’s no net selection even though they’re diseases]

Related: a reader points out this paper on “The Social Construction Of Hungarian Genius”.

PNAS has a good (albeit kind of silly) article on claims that scientific progress has slowed.

This month in insane Twitter drama, for people who have previously made the good choice not to follow insane Twitter drama but want to walk back on it for some reason: will Sam Kriss publicly denounce Zionism? (1, 2, 3). What happens when Jeet Heer tweets “Bernie would have won”? Is Joan Walsh un-woke on Palestine? And apparently there is some kind of Joan Walsh/Katie Halper feud. I realize this sort of stuff seems petty, but it was really helpful in getting me to understand why everyone hates each other, and helped convince me that a lot of things I thought were silly arguments against straw men are actually important arguments against a large contingent of (depressingly) real people.

Parcel sorting facilities in China.

Noah Smith: The US has forgotten how to do infrastructure.

That story about how Gavrilio Princip failed to assassinate Archduke Ferdinand, then went to get a sandwich, then ran into the Archduke just outside the sandwich shop and assassinated him anyway? The Smithsonian says it’s probably false.

Did you know: the ancient Egyptian language of hieroglyphs and Pharaohs survived into modern times as the Coptic language and is still the liturgical language of Coptic churches today. Also, English words derived from Egyptian include “adobe” and “oasis”.

New study finds that growth mindset is not associated with scholastic aptitude in a large sample of university applicants. Particularly excited about this one because an author said that my blog posts about growth mindset inspired the study. I’m honored to have been able to help the progress of science!

Rocket pioneer Wernher von Braun wrote a 1952 sci-fi story about the colonization of Mars, in which the Martian government was led by a President called “the Elon”.

Internet payment processors and payment regulations are terrible and destroying the nootropics industry and probably a bunch of other industries I know less about, part 459401.

ABC News reported that a meat company’s product was disgusting low-quality “pink slime”, and people stopped buying from that meat company. Now the company is suing ABC for $5.7 billion in a lawsuit with the potential to have chilling effects on journalism in general.

I think the Byzantine Empire had the coolest-sounding titles of any civilization, including Grand Logothete and Megaduke.

A deadly fire in an apartment building in London gains an extra layer of horror in the context of this blog post by apartment residents predicting that there is definitely going to be a deadly fire there soon due to apartment management and local government incompetence.

Contra recent thinkpieces about how polls don’t work and psephology is a pseudoscience, on an aggregate level the probabilities from prediction markets have been impressively accurate.

Phone companies were ready to deploy cellular phones since the 1940s – the reason we didn’t get them until the ’80s was government regulators refusing to give them the spectrum space for political reasons.

There’s wide state-by-state variation on the legality of shooting Bigfoot, with Washington calling it a felony and Texas calling it acceptable given that it’s technically “an invasive species”. (h/t Tumblr)

A swarm of 20,000 bees recently descended upon Vox Media’s Manhattan offices, leading to articles like The Swarm Of Bees Outside Vox Media, Explained.

Posted in Uncategorized | Tagged | 1,007 Comments

What Is Depression, Anyway?: The Synapse Hypothesis

I.

The problem with depression research isn’t that we don’t have any leads on what causes depression. It’s that we have so many leads on what causes depression that we don’t know what to do with all of them. For example:

1. Life adversity, like getting fired or breaking up with a partner, can make people depressed. The biological correlate of this seems to be the hypothalamic–pituitary–adrenal axis (HPA), where your brain tells your adrenal glands to produce glucocorticoid stress hormones like cortisol and this does something to your brain that increases the risk of depression.

2. Inflammation and immune overactivity can make people depressed. The classic examples of this are cancer-related depression (which exceeds what you would expect just from cancer being stressful) and depression induced by administration of the immunomodulator interferon-a. Antiinflammatory drugs have a small but clinically relevant antidepressant effect. Some of the relevant chemicals here seem to be TNF-A and IL-1; these do something to your brain that increases the risk of depression.

3. Serotonin and other monoamines seem to be involved. Most existing antidepressants, like SSRIs and MAOIs, seem to work by increasing monoamine levels. There are some conditions which affect monoamine levels and also increase risk of depression, though it’s nothing like a perfect correlation.

4. The glutamate system (eg NMDA and AMPA receptors) seem to be involved. Ketamine acts on both of these receptors in different ways, and one of those actions is the source of its rapid and unprecedented antidepressant effects.

5. There’s some kind of important link between depression and folate balance. Various folate-related chemicals (eg l-methylfolate and s-adenosylmethionine) are effective antidepressants. Some studies show that people with depression sometimes have disrupted folate cycles, for example elevated homocysteine levels.

6. Electroconvulsive therapy (“shock therapy”) is very effective at treating depression if it induces a seizure in the patient, so the increased activity from seizures must be helpful somehow.

So if we wanted to know what depression really was, it might be promising to look for some process that seems to match depressive symptoms and affects/is affected by life adversity, inflammation, monoamines, glutamate, folate, and electricity.

Recently some people think they’ve found one. According to Duman’s Neurobiology of Stress, Depression, and Rapid Acting Antidepressants, it’s decreased synaptogenesis, and it’s regulated by a protein complex called mTORC1.

Neurons communicate with other neurons through branches called dendrites and connections called synapses. Healthy neurons often create new dendrites and synapses to expand their network of connections and adjust to new information. The process of making new synapses is called “synaptogenesis”, and it’s common throughout the adult brain.

As mentioned above, depressed people have decreased volume in some brain areas. But in postmortem studies, they don’t actually have fewer cells in those areas. So it looks like maybe these neurons just have less synaptogenesis going on.

Synaptogenesis is partly controlled by a protein complex called mechanistic target of rapamycin complex 1 (mTORC1 to its friends). Like every other protein, mTORC is controlled by a giant mess of receptors and second messengers and intracellular signals with names like VDCC and GSK3.

People try to make this seem simple by displaying it as a system of billiard balls and tubes in a cute cartoon, but don’t be fooled – no human being has ever remembered any of it for more than two seconds.

The factors that affect synaptogenesis and mTORC are many of the same factors that affect depression. Let me count the ways:

1. Life adversity causes chronic stress, biologically represented by upregulation of the HPA axis and increased corticosteroid production. A 2008 study finds that rats who are subjected to chronic stress develop atrophy of dendrites in their prefrontal cortex. Administering glucocorticoids directly mimicked some of these effects, suggesting that stress is a whole cocktail of things including glucocorticoids and other things. When humans take glucocorticoids (they’re a useful medicine for various diseases) they tend to develop hippocampal atrophy and “simplification of dendrites” there, which I think is the same as decreased synaptogenesis. They also tend to get depressed – in some studies of Cushing’s Syndrome (the medical name for the collection of bad things that happen when you take too much glucocorticoid medication), up to 90% of patients are depressed.

2. I didn’t find the linked paper’s attempt to link inflammation to synaptogenesis very convincing, but it looks like there’s a little bit of research that has found that systemic inflammation decreases synaptogenesis. “Morphometric analysis of dendritic spines identified a period of vulnerability, manifested as a decrease in [dendritic] spine density in response to inflammation. The density of presynaptic excitatory terminals was similarly affected. When the systemic inflammation was extended from 24h to 8 days, the negative effects on the excitatory terminals were more pronounced and suggested a reduced excitatory drive.” This seems pretty relevant.

3. Everyone used to think that traditional antidepressants like SSRIs worked by increasing serotonin (and so by extension depression must have something to do with low serotonin levels). But SSRIs increase serotonin very quickly (within hours) yet take months to work. Something longer-term must happen when serotonin levels have been increased for long enough. That something has now been pretty conclusively identified as an increase in brain-derived neurotrophic factor (BDNF) – although I can’t find any good explanation of why increased serotonin should cause increased BDNF after a month. BDNF is a nerve growth factor – its main action is activating mTORC and telling nerve cells to grow more dendrites and synapses. And it’s most active in the cortex and hippocampus.

4. Ketamine affects the brain by either blocking NMDA receptors (boring traditional explanation), activating AMPA receptors (exciting new explanation), or possibly both (wishy-washy neoliberal compromise explanation). Duman et al are kind of ambiguous about which explanation they accept, but I think they present a theory where NMDA blockade causes AMPA activation, or something, which I’d never heard before. In any case, they present ample evidence that AMPA rapidly affects BDMF and dendritogenesis – for example, Positive AMPA Receptor Modulation Rapidly Stimulates BDNF Release And Increases Dendritic MRNA Translation. The “rapidly” part is important – the surprising thing about ketamine is how quickly it works compared to other antidepressants, so it’s exciting to find a theory that predicts this should happen.

5. I haven’t seen much attempt to fit folate into this theory, which is a shame. A quick Google search brings up a few people talking about how folate deficiency decreases neurogenesis in the hippocampus, which is sort of related.

6. Studies show that ECT increases BDNF levels and increases hippocampal volume, though I’m not sure exactly how or why giving someone a seizure should do that.

So the synapse hypothesis can unify at least five of the six lines of research into the causes of depression.

II.

My remaining skepticism is mostly based on a worry that anyone can do this with anything. The body is so interconnected, and there’s so much bad biology research out there, that I worry that if I said that the real cause of depression was, uh, thickness of the blood, I could find some way that all of those lines of research above affected blood thickness.

A quick demonstration: glucocorticoids can cause thicker blood, inflammation can cause thicker blood, SSRIs cause thinner blood, folate causes thinner blood. Huh, actually that’s kind of creepy.

My point isn’t that the (very respectable) academic research on depression is anywhere near this silly. It’s just to explain why I can hear a theory that seems to explain everything beautifully and my only reaction is “Eh, sounds like it has potential, let’s see what happens.”

Here are some of the things that confuse me, or that I hope get researched more in the future:

1. Why should decreased synaptogenesis cause depression, of all things? If you asked me, a non-neuroscientist, to guess what happens if the brain can’t create new synapses very well and loses hippocampal volume, I would say “your memory gets worse and you stop being able to learn new things”. But this doesn’t really happen in depression – even the subset of depressed people who get cognitive problems usually just have “pseudo-dementia” – they’re too depressed to put any effort into answering questions or doing intelligence tests. Why should decreased synaptogenesis in the hippocampus and prefrontal cortex cause poor mood, tiredness, and even suicidality? All that the Duman et al paper has to say about this is:

This reduction in dendrite complexity and synaptic connections could contribute to the decreased volume of PFC and hippocampus observed in depressed patients. Moreover, loss of synaptic connections could contribute to a functional disconnection and loss of normal control of mood and emotion in depression (Fig. 1). In particular, the medial PFC exerts top down control over other brain regions that regulate emotion and mood, most notably the amygdala, and loss of synaptic connections from PFC to this and other brain regions could thereby result in more labile mood and emotion, as well as cognitive deficits.

…which sounds more like an IOU for a theory than anything really fleshed out.

2. Why can’t we just give people BDNF for depression? I’ve been looking into this and it seems like the answer is something like “this works great if you cut open someone’s skull and inject it directly into their brain, but most people aren’t up for it” (the relevant studies were done in rats). But why can’t it be given peripherally? Some studies suggest it’s stable on injection and crosses the blood-brain barrier. Some people tried this in mice and got modest results, but why aren’t people looking into it more?

3. Why does the body have so many “decrease synaptogenesis” knobs? That is, why go through the trouble to evolve all these chemicals and systems whose job is to tell your brain to decrease synapse formation so much that you end up depressed? Is there some huge problem with having too much synapse formation which the brain is desperately trying to avoid? For that matter, what is it like to have too much synapse formation? If it’s the opposite of depression, it sounds kind of fun. If I got someone to open up my skull and inject a lot of BDNF, could I be really happy and energetic all the time? How come all the good stuff is always reserved for rats?

4. Why is depression an episodic disease? That is, how come so many people get depressed for no reason, stay depressed for a few months to a few years, and then get better – only to relapse back into depression a few years later? If people get depressed because of some life stressor like a divorce, how come they don’t get un-depressed once the life stressor goes away? Is depression some kind of attractor state? If so, why?

5. Why doesn’t rapamycin cause depression? Remember, mTORC is “mechanistic target of rapamycin”, so named because the drug rapamycin inhibits it. But we give people rapamycin for various things all the time, and depression isn’t really known as a major side effect (even though IIRC it crosses the blood-brain barrier). If depression is really under the immediate control of mTORC, rapamycin should be the most depressive thing. Instead it’s not obviously depressive at all.

6. How does bipolar disorder fit into all of this? Is mania the answer to my “what is it like to have too many synapses?” question from point (3)? If so, why do some people go back and forth between that and depression?

A lot of these questions could be answered in one stroke if we had a good evolutionary theory of depression. I’m skeptical that this exists – depression just seems too fitness-decreasing, and the various just-so stories people have come up with for why it might increase fitness in certain weird situations seem a little too convoluted. So it’s not that I’m expecting some sort of evolutionary story to work out. Just noticing that, even if the synapse theory of pathophysiology turns out to be right, there’s still a lot more that needs to be explained.

Posted in Uncategorized | Tagged | 145 Comments

SSC Journal Club: AI Timelines

I.

A few years ago, Muller and Bostrom et al surveyed AI researchers to assess their opinion on AI progress and superintelligence. Since then, deep learning took off, AlphaGo beat human Go champions, and the field has generally progressed. I’ve been waiting for a new survey for a while, and now we have one.

Grace et al (New Scientist article, paper, see also the post on the author’s blog AI Impacts) surveyed 1634 experts at major AI conferences and received 352 responses. Unlike Bostrom’s survey, this didn’t oversample experts at weird futurist conferences and seems to be a pretty good cross-section of mainstream opinion in the field. What did they think?

Well, a lot of different things.

The headline result: the researchers asked experts for their probabilities that we would get AI that was “able to accomplish every task better and more cheaply than human workers”. The experts thought on average there was a 50% chance of this happening by 2062 – and a 10% chance of it happening by 2026!

But on its own this is a bit misleading. They also asked by what year “for any occupation, machines could be built to carry out the task better and more cheaply than human workers”. The experts thought on average that there was a 50% chance of this happening by 2139, and a 20% chance of it happening by 2037.

As the authors point out, these two questions are basically the same – they were put in just to test if there was any framing effect. The framing effect was apparently strong enough to shift the median date of strong human-level AI from 2062 to 2139. This makes it hard to argue AI experts actually have a strong opinion on this.

Also, these averages are deceptive. Several experts thought there was basically a 100% chance of strong AI by 2035; others thought there was only a 20% chance or less by 2100. This is less “AI experts have spoken and it will happen in 2062” and more “AI experts have spoken, and everything they say contradicts each other and quite often themselves”.

This does convey more than zero information. It conveys the information that AI researchers are really unsure. I can’t tell you how many people I’ve heard say “there’s no serious AI researcher who thinks there’s any chance of human-level intelligence before 2050”. Well actually, there are a few dozen conference-paper-presenting experts who think there’s a one hundred percent chance of human-level AI before that year. I don’t know what drugs they’re on, but they exist. The moral of the story is: be less certain about this kind of thing.

II.

The next thing we can take from this paper is a timeline of what will happen when. The authors give a bunch of different tasks, jobs, and milestones, and ask the researchers when AI will be able to complete them. Average answers range from nearly fifty years off (for machines being able to do original high-level mathematical research) to only three years away (for machines achieving the venerable accomplishment of being able to outperform humans at Angry Birds). Along the way they’ll beat humans at poker (four years), writing high school essays (ten years), be able to outrun humans in a 5K foot race (12 years), and write a New York Times bestseller (26 years). What do these AI researchers think is the hardest and most quintessentially human of the tasks listed, the one robots will have the most trouble doing because of its Olympian intellectual requirements? That’s right – AI research (80 years).

I make fun of this, but it’s actually interesting to think about. Might the AI researchers have put their own job last not because of an inflated sense of their own importance, but because they engage with it every day in Near Mode? That is, because they imagine writing a New York Times bestseller as “something something pen paper be good with words okay done” whereas they understand the complexity of AI research and how excruciatingly hard it would be to automate away every piece of what they do?

Also, since they rated AI research (80 years) as the hardest of all occupations, what do they mean when they say that “full automation of all human jobs” is 125 years away? Some other job not on the list that will take 40 years longer than AI research? Or just a combination of framing effects and not understanding the question?

(it’s also unclear to what extent they believe that automating AI research will lead to a feedback loop and subsequent hard takeoff to superintelligence. This kind of theory would fit with it being the last job to be automated, but not with it taking another forty years before an unspecified age of full automation.)

III.

The last part is the most interesting for me: what do AI researchers believe about risk from superintelligence?

This is very different from the earlier questions about timelines. It’s possible to believe that AI will come very soon but be perfectly safe. And it’s possible to believe that AI is a long time away but we really need to start preparing now, or else. A lot of popular accounts collapse these two things together, “oh, you’re worried about AI, but that’s dumb because there’s no way it’s going to happen anytime soon”, but past research has shown that short timelines and high risk assessment are only modestly correlated. This survey asked about both separately.

There were a couple of different questions trying to get at this, but it looks like the most direct one was “does Stuart Russell’s argument for why highly advanced AI might pose a risk, point at an important problem?”. You can see the exact version of his argument quoted in the survey on the AI Impacts page, but it’s basically the standard Bostrom/Yudkowsky argument for why AIs may end up with extreme values contrary to our own, framed in a very normal-sounding and non-threatening way. According to the experts, this was:

No, not a real problem: 11%
No, not an important problem: 19%
Yes, a moderately important problem: 31%
Yes, an important problem: 34%
Yes, among the most important problems in the field: 5%

70% of AI experts agree with the basic argument that there’s a risk from poorly-goal-aligned AI. But very few believe it’s among “the most important problems in the field”. This is pretty surprising; if there’s a good chance AI could be hostile to humans, shouldn’t that automatically be pretty high on the priority list?

The next question might help explain this: “Value of working on this problem now, compared to other problems in the field?”

Much less valuable: 22%
Less valuable: 41%
As valuable as other problems: 28%
More valuable: 7%
Much more valuable: 1.4%

So charitably, the answer to this question was coloring the answer to the previous one: AI researchers believe it’s plausible that there could be major problems with machine goal alignment, they just don’t think that there’s too much point in working on it now.

One more question here: “Chance intelligence explosion argument is broadly correct?”

Quite likely (81-100% chance): 12%
Likely (61-80% chance): 17%
About even (41-60% chance): 21%
Unlikely (21-40% chance): 24%
Quite unlikely (0-20% chance): 26%

Splitting the 41-60% bin in two, we might estimate that about 40% of AI researchers think the hypothesis is more likely than not.

Take the big picture here, and I worry there’s sort of a discrepancy.

50% of experts think there’s at least a ten percent chance of above-human-level AI coming within the next ten years.

And 40% of experts think that there’s a better-than-even chance that, once we get above-human level AI, it will “explode” to suddenly become vastly more intelligent than humans.

And 70% of experts think that Stuart Russell makes a pretty good point when he says that without a lot of research into AI goal alignment, AIs will probably have their goals so misaligned with humans that they could become dangerous and hostile.

I don’t have the raw individual-level data, so I can’t prove that these aren’t all anti-correlated in some perverse way that’s the opposite of the direction I would expect. But if we assume they’re not, and just naively multiply the probabilities together for a rough estimate, that suggests that about 14% of experts believe that all three of these things: that AI might be soon, superintelligent, and hostile.

Yet only a third of these – 5% – think this is “among the most important problems in the field”. Only a tenth – 1.4% – think it’s “much more valuable” than other things they could be working on.

IV.

How have things changed since Muller and Bostrom’s survey in 2012?

The short answer is “confusingly”. Since almost everyone agrees that AI progress in the past five years has been much faster than expected, we would expect experts to have faster timelines – ie expect AI to be closer now than they did then. But Bostrom’s sample predicted human-level AI in 2040 (median) or 2081 (mean). Grace et al don’t give clear means or medians, preferring some complicated statistical construct which isn’t exactly similar to either of these. But their dates – 2062 by one framing, 2139 by another – at least seem potentially a little bit later.

Some of this may have to do with a subtle difference in how they asked their question:

Bostrom: “Define a high-level machine intelligence as one that can carry out most human professions as well as a typical human…”

Grace: “High-level machine intelligence is achieved when unaided machines can accomplish every task better and more cheaply than human workers.”

Bostrom wanted it equal to humans; Grace wants it better. Bostrom wanted “most professions”, Grace wants “every task”. It makes sense that experts would predict longer timescales for meeting Grace’s standards.

But as we saw before, expecting AI experts to make sense might be giving them too much credit. A more likely possibility: Bostrom’s sample included people from wackier subbranches of AI research, like a conference on Philosophy of AI and one on Artificial General Intelligence; Grace’s sample was more mainstream. The most mainstream part of Bostrom’s sample, a list of top 100 AI researchers, had an estimate a bit closer to Grace’s (2050).

We can also compare the two samples on belief in an intelligence explosion. Bostrom asked how likely it was that AI went from human-level to “greatly surpassing” human level within two years. The median was 10%; the mean was 19%. The median of top AI researchers not involved in wacky conferences was 5%.

Grace asked the same question, with much the same results: a median 10% probability. I have no idea why this question – which details what an “intelligence explosion” would entail – was so much less popular than the one that used the words “intelligence explosion” (remember, 40% of experts agreed that “the intelligence explosion argument is broadly correct”). Maybe researchers believe it’s a logically sound argument and worth considering but in the end it’s not going to happen – or maybe they don’t actually know what “intelligence explosion” means.

Finally, Bostrom and Grace both asked experts’ predictions for whether the final impact of AI would be good or bad. Bostrom’s full sample (top 100 subgroup in parentheses) was:

Extremely good: 24% (20)
On balance good: 28% (40)
More or less neutral: 17% (19)
On balance bad: 13% (13)
Extremely bad – existential catastrophe: 18% (8)

Grace’s results for the same question:

Extremely good: 20%
On balance good: 25%
More or less neutral: 40%
On balance bad: 10%
Extremely bad – human extinction: 5%

Grace’s data looks pretty much the same as the TOP100 subset of Bostrom’s data, which makes sense since both are prestigious non-wacky AI researchers.

V.

A final question: “How much should society prioritize AI safety research”?

Much less: 5%
Less: 6%
About the same: 41%
More: 35%
Much more: 12%

People who say that real AI researchers don’t believe in safety research are now just empirically wrong. I can’t yet say that most of them want more such research – it’s only 47% on this survey. But next survey AI will be a little bit more advanced, people will have thought it over a little bit more, and maybe we’ll break the 50% mark.

But we’re not there yet.

I think a good summary of this paper would be that large-minorities-to-small-majorities of AI experts agree with the arguments around AI risk and think they’re worth investigating further. But only a very small minority of experts consider it an emergency or think it’s really important right now.

You could tell an optimistic story here – “experts agree that things will probably be okay, everyone can calm down”.

You can also tell a more pessimistic story. Experts agree with a lot of the claims and arguments that suggest reason for concern. It’s just that, having granted them, they’re not actually concerned.

This seems like a pretty common problem in philosophy. “Do you believe it’s more important that poor people have basic necessities of life than that you have lots of luxury goods?” “Yeah” “And do you believe that the money you’re currently spending on luxury goods right now could instead be spent on charity that would help poor people get life necessities?” “Yeah.” “Then shouldn’t you stop buying luxury goods and instead give all your extra money beyond what you need to live to charity?” “Hey, what? Nobody does that! That would be a lot of work and make me look really weird!”

How many of the experts in this survey are victims of the same problem? “Do you believe powerful AI is coming soon?” “Yeah.” “Do you believe it could be really dangerous?” “Yeah.” “Then shouldn’t you worry about this?” “Hey, what? Nobody does that! That would be a lot of work and make me look really weird!”

I don’t know. But I’m encouraged to see people are even taking the arguments seriously. And I’m encouraged that researchers are finally giving us good data on this. Thanks to the authors of this study for being so diligent, helpful, intelligent, wonderful, and (of course) sexy.

(I might have forgotten to mention that the lead author is my girlfriend. But that’s not biasing my praise above in any way.)

Posted in Uncategorized | Tagged , | 283 Comments

Is Pharma Research Worse Than Chance?

[Epistemic status: very speculative]

The two most exciting developments in psychopharmacology in the 21st century so far have been ketamine for depression and MDMA for PTSD.

Unlike other antidepressants, which work intermittently over a space of weeks, ketamine can cause near-instant remission of depression with a single infusion – which lasts a week or two and can be repeated if needed. Ketamine use may be successful in 50-70% of patients who have failed treatment with conventional antidepressants. Ketamine treatment has some issues right now, but the race is on to create an oral non-hallucinogenic version which could be the next big blockbuster drug and revolutionize depression treatment.

MDMA (“Ecstasy”) is undergoing FDA Phase 3 clinical trials as a treatment for PTSD. Preliminary research has been small and underpowered, but suggests response rates up to 80% and effect sizes greater than 1 in this otherwise-hard-to-treat condition. None of this is on really firm footing – that’ll have to wait for the Phase 3. But signs are looking very good.

I say these are the two most exciting developments mostly because no other developments have been exciting. In terms of normal psychiatric drugs, the best that the 21st century has given us has probably been pimavanserin and aripiprazole, modest updates to the standard atypical antipsychotic model. These drugs are probably a bit better than existing ones for the people who need them (especially pimavenserin for psychosis in Parkinson’s) but they don’t revolutionize the treatment of any condition and nobody ever claimed that they did. And most drugs aren’t even at this level – they’re new members of well-worn classes with slightly different side effect profiles. The landscape was so quiet that ketamine came in like a bolt from the blue, and MDMA is set to do the same in a couple of years when the trial results come out.

(if I’m wrong, and history decides these two drugs weren’t the biggest developments, the most likely failure mode is that psilocybin turned out to be more important than MDMA)

There’s a morality tale to be told here about how the War on Drugs choked off vital research on some of the most powerful psychiatric compounds and cost us fifty years in exploring these effects and treating patients. I agree with this morality tale as far as it goes, but I also think there’s another, broader morality tale beneath it.

Suppose that neither ketamine nor MDMA were illegal drugs. Ketamine was just used as an anaesthetic. MDMA was just used as a chemical intermediate in producing haemostatic drugs, its original purpose. Now the story is that, fifty years later, we learn that this anaesthetic and this haemostatic turn out to have incredibly powerful psychiatric effects. What’s our narrative now?

For me it’s about the weird inability of intentional psychopharmaceutical research to discover anything as good as things random druggies use to get high.

For decades, pharmaceutical companies have been coming out with relatively lackluster mental health offerings – aripiprazole, pimavanserin, and all the rest. And when asked why, they answer that mental health is hard, the brain is the most complicated organ in the known universe, we shouldn’t expect there to be great cures with few side effects for psychiatric diseases, and if there were we certainly shouldn’t expect them to be easy to find.

And this would make sense except in the context of ketamine and MDMA. Here are some random chemicals that affect the brain in some random way, which people were using mostly because they felt good at raves, and huh, they seem to treat psychiatric diseases much better than anything produced by some of the smartest people in the world working for decades on ways to treat psychiatric diseases. Why should that be?

One could argue it’s all about numbers vs. base rates. There are way more chemicals synthesized each year by people who aren’t looking for psychiatric drugs than by people who are. Even if the people who are looking for drugs are a thousand times more likely to find them, the people-who-aren’t-looking can still overwhelm them with sheer numerical advantage. And maybe when a psychiatric drug is discovered by people who weren’t looking for it, what this looks like is a few random people trying it, noticing it feels good, and turning it into a drug of abuse.

And I’m sure this is part of the story. But that just passes the buck to the next question. Abusers take the vast flood of possible chemicals and select the ones they think will feel good at raves. Psychopharmacologists take the vast flood of possible chemicals and select the ones they think will treat mental illnesses. How come the abusers’ selection process is better at picking out promising mental health treatments?

Here’s one hypothesis: at the highest level, the brain doesn’t have that many variables to affect, or all the variables are connected. If you smack the brain really really hard in some direction or other, you will probably treat some psychiatric disease. Drugs of abuse are ones that smack the brain really hard in some direction or other. They do something. So find the psychiatric illness that’s treated by smacking the brain in that direction, and you’re good.

(in fact, the most effective existing treatment for depression is electroconvulsive therapy – ie giving the brain a big electric shock. This is maybe the crudest, most literally “smack the brain really hard” treatment out there, but it sure does work)

Actual carefully-researched psychiatric drugs are exquisitely selected for having few side effects. The goal is something like an SSRI – mild stomach discomfort, some problems having sex, but overall you can be on them forever and barely notice their existence. In the grand scheme of things their side effects are tiny – in most placebo-controlled studies, people have a really hard time telling whether they’re in the experimental or the placebo group.

Nobody has a hard time telling whether they’re in the experimental or placebo group of a trial of high-dose MDMA. I think this might be the difference. If you go for large effects – even if you don’t really care what direction the effect is in – you’ll get them. And if you go for small, barely perceptible effects, then you’ll get those too. The dream of the magic bullet – the drug that treats exactly what it’s supposed to treat but otherwise has no effect at all on you – is just a dream. The closest you can come is something with miniscule side effects but a barely-less-miniscule treatment effect.

But given that we’re all very excited to learn about ketamine and MDMA, and given that if their original promise survives further testing we will consider them great discoveries (and given that ECT was also a great and productive discovery) it suggests we chose the wrong part of the tradeoff curve. Or at least it suggests a different way of framing that tradeoff curve. A drug that makes you feel extreme side effects for a few hours – but also has very strong and lasting treatment effects – is better than a drug with few side effects and weaker treatment effects. That suggests a new direction pharmaceutical companies might take: look for the chemicals that have the strongest and wackiest effects on the human mind. Then see if any of them also treat some disease.

I think this is impossible with current incentives. There’s too little risk-tolerance at every stage in the system. But if everyone rallied around the idea, it might be that trying the top hundred craziest things Alexander Shulgin dreamed up on whatever your rat model is would be orders of magnitude more productive than whatever people are doing now.

Or it might not be. I can also think of a counterargument to the theory above, which is that our current best model of ketamine suggests it’s a non-psychoactive metabolite that has most of the useful antidepressant effect. In fact, a lot of people think that one form of ketamine is hallucinogenic (and extremely effective against chronic pain) and another form (or its metabolite) is the antidepressant. I’m a little suspicious trying to calculate the odds of a single chemical having two forms, one of which is a really exciting analgesic, and the other of which is a really exciting antidepressant, by two different mechanisms. It sounds too much like finding some new chemical compound whose solid form is a room-temperature superconductor, and whose liquid form catalyzes cold fusion, by two totally different mechanisms. It seems a little too lucky (see here for some ketamine skepticism, and here for my response). But if it were true, it means that ketamine’s psychoactive effects were a red herring in helping us discover it as an antidepressant, even though they were a very effective red herring.

OT77: Opium Thread

This is the bi-weekly visible open thread. Post about anything you want, ask random questions, whatever. You can also talk at the SSC subreddit, the SSC Discord server, or the Cafe Chesscourt forum.

1. Comments of the week: Mazirian on Ashkenazi demographic expansion, Majuscule on history of the Jews in Hungary, and B (at Information Processing) on the role of assimilation. And from the subreddit, an experimental philosophy question and very clever response.

2. Thanks to an SSC reader who will remain unnamed until I remember to ask him if he wants his real name on the Internet, we have an Esperanto copy of Laszlo Polgar’s Bring Up Genius (as mentioned here) and an offer to translate it into English for $1200. If you want to help pay, donate at the GoFundMe page [EDIT: Campaign is complete, thank you!] Legal issues permitting, I’ll try to post the translation here once I have it.

3. A new ad on the sidebar, Meditation For Atheists, an audio course by an SSC reader interested in non-religious meditation practices. If you’re interested in advertising on SSC, you can find more information here.

Posted in Uncategorized | Tagged | 1,065 Comments

Hungarian Education III: Mastering The Core Teachings Of The Budapestians

[Previously in series: The Atomic Bomb Considered As A Hungarian High School Science Fair Project, Four Nobel Truths]

I.

Someone summed up my previous post as “Hungarian education isn’t magic”. I would amend that to read “Hungarian education isn’t systemically magic”. As far as I know, there’s only one Hungarian educator with magic powers, and (like all good wizards) his secrets are maddeningly hard to find.

Laszlo Polgar studied intelligence in university, and decided he had discovered the basic principles behind raising any child to be a genius. He wrote a book called Bring Up Genius and recruited an interested woman to marry him so they could test his philosophy by raising children together. He said a bunch of stuff on how ‘natural talent’ was meaningless and so any child could become a prodigy with the right upbringing.

This is normally the point where I’d start making fun of him. Except that when he trained his three daughters in chess, they became the 1st, 2nd, and 6th best female chess players in the world, gaining honors like “youngest grandmaster ever” and “greatest female chess player of all time”. Also they spoke seven languages, including Esperanto.

Their immense success suggests that education can have a major effect even on such traditional genius-requiring domains as chess ability. How can we reconcile that with the rest of our picture of the world, and how obsessed should we be with getting a copy of Laszlo Polgar’s book?

II.

Let’s get this out of the way first: the Polgar sisters were probably genetically really smart. The whole family was Hungarian Jews, a group with a great track record. Their mother and father were both well-educated teachers interested in stuff like developmental psychology. They had every possible biological advantage and I’m sure that helped.

J Levitt proposes an equation to estimate a chess player’s IQ from their chess score. It suggests that chess grandmasters probably have IQs above 160. Plugging the Polgar sisters’ chess scores into his equation, I get IQs in the range of 150, 160, and 170 for the three sisters.

This is biologically impossible. Even if both Polgar parents were 170-IQ themselves, regression to the mean predicts that their children would have IQs around 140 to 150. It’s mathematically possible for there to be an IQ that predicts you would have three children of 150, 160, and 170, but I doubt any living people have it, and even if they did there’s no way they would marry somebody else equally gifted.

[EDIT: Thanks to a few people who pointed out some problems with my math here (1, 2, 3). I still think that having three supergenius-IQ kids when you and your spouse show no signs of being a supergenius yourself (Laszlo Polgar’s daughters could beat him at chess by the time they were 8) is pretty unlikely, but I admit not impossible. I still think arguing about this is unnecessary thanks to the points below.]

On the other hand, I’m not sure Levitt’s right. Chess champion Gary Kasparov actually sat and took an IQ test for the magazine Der Spiegel, and his IQ was 135. That’s not bad – it’s top 1% of the population – but it’s not amazing either.

This is what we should expect given the correlation of about r = 0.24 between IQ and chess ability (see also this analysis, although I disagree with the details). And the contrary claims – like the one that Bobby Fischer’s IQ was in the 180s – are less well-sourced (although Fischer was the son of a Hungarian-Jewish mathematician, so who knows?).

If it were possible to be a chess world champion with an IQ of 135, then maybe it’s possible to be a “mere” grandmaster with IQs in the high 120s and low 130s. And it’s just barely plausible that some sufficiently smart people might have three kids who all have IQs in the high 120s and low 130s.

But this just passes the buck on the mystery. 2% of people have IQs in the high 120s or low 130s, but 2% of people aren’t the top-ranked female chess player in the world. The Polgar sisters’ IQs might have been a permissive factor in allowing them to excel, but it didn’t necessitate it. So what’s going on there?

III.

“Practice” seems like an obvious part of the picture. Malcolm Gladwell uses the Polgars as poster children for his famous ‘10,000 hours of practice makes you an expert at anything’ rule. The Polgars had 50,000 hours of chess practice each by the time they were adults, presumably enough to make them quintuple-experts.

Robert Howard has a paper Does High-Level Performance Depend On Practice Alone? Debunking The Polgar Sisters Case in which he argues against the strong version of Gladwell’s thesis. He points out that there are many chess masters who have practiced much less than the Polgar sisters but are better than they are. He also points out that even though the sisters themselves have all practiced similar amounts, youngest sister Judit is clearly better than the other two in a way that practice alone cannot explain.

I don’t know if the case he’s arguing against – that practice is literally everything and it’s impossible for anything else to factor in – is a straw man or not. But it seems more important to consider a less silly argument – that practice is one of many factors, and that enough of it can make up for a lack of the others. This seems potentially true. This study showing that amount of practice only explains 12% of the variance in skill level at various tasks, and is often summarized as “practice doesn’t matter much”. But it finds practice matters more (25% of the variance) in unchanging games with clear fixed rules, and uses chess as an example.

So suppose that the Polgar sisters are genetically smart, but maybe not as high up there as some other chess masters. We would expect them to need much more practice to achieve a level of proficiency similar to those chess masters, and indeed that seems like what happens.

(all of this is confounded by them being women and almost all the other equally-good chess masters being men. It’s unclear if the Polgars deserve extra points for overcoming whatever factor usually keeps women out of the highest levels of chess.)

But I’m actually still not sure this suffices as an explanation. According to Wikipedia:

Polgár began teaching his eldest daughter, Susan, to play chess when she was four years old. Six months later, Susan toddled into Budapest’s smoke-filled chess club,” which was crowded with elderly men, and proceeded to beat the veteran players.

The study linked above suggests that Susan practiced 48 hours a week. During those six months, she would have accumulated about 1200 hours of practice. Suppose the elderly Budapest chess players practiced only one hour a week, but had been doing so for the last twenty-five years. They would have more practice than Susan – plus the advantage of having older, more developed brains. So why did she beat them so easily?

Maybe there’s a time-decay factor for practice? That is, maybe Susan had been practicing intensively, so she got a lot of chances to link it all together as she was learning, and also it was fresh in her mind when she went to the club to go play? I’m not sure. If some of those veterans had been playing more than one hour a week (and surely the sort of people who frequent Budapest chess clubs do) then her advantage seems too implausible to be due to freshness-of-material alone.

IV.

That leaves two possibilities.

First, Susan could have benefitted from some form of malleability. A lot of people claim there’s a “developmental window” during which children have a unique ability to learn language. If cats see only vertical stripes for the first few weeks of their lives, they never learn to see in horizontal. Maybe if you teach your kid high-level chess at age 4, they’ll be able to recruit systems that adults could never manage, or reorganize the fundamental structure of their brain to conform to chess better, or something like that.

Second, Polgar might actually have some really good educational methods besides just “start early and have a lot of practice”. I assume this is true, but I’m having a lot of trouble finding them. Shockingly, Polgar’s book Bring Up Genius is out of print and totally unavailable anywhere – I guess the book-reading community heard that someone wrote down a way to reliably turn any child into a genius which had a great real-world track record of success, and collectively decided “Nah, better read Fifty Shades Of Grey instead”. I’m not sure at what point I should start positing a conspiracy of suppression, or whether that would be better or worse than the alternative.

The book seems to possibly be available in Hungarian under the title Nevelj zsenit!, but I can’t tell for sure and a lot of the Hungarian sites suggest it’s out of print even in that language. There may have been a recent republication in Esperanto called Eduku geniulon!, but I can’t find that one either. If anybody knows where to find this book and wants to send it to me, I will figure out some way to translate it and review it. I’d also be willing to pay for costs and even pay extra for your time if it helps. Come on, Esperanto-speakers! This is the only chance you’ll ever have to be useful!

V.

One thing I know without reading the book: Polgar says that his method should work to create geniuses in any field, not just chess. He said he chose chess kind of on the whim of his eldest daughter. From Wikipedia:

Polgár and his wife considered various possible subjects in which to drill their children, “including mathematics and foreign languages,” but they settled on chess. “We could do the same thing with any subject, if you start early, spend lots of time and give great love to that one subject,” Klara later explained. “But we chose chess. Chess is very objective and easy to measure.” Susan described chess as having been her own choice: “Yes, he could have put us in any field, but it was I who chose chess as a four-year-old…. I liked the chessmen; they were toys for me.”

It’s disappointing that he decided to stick with chess for his other two daughters. The study linked above suggests that chess is unusually amenable to practice. What would have happened if he’d tried to train his kids in art? In mathematics? In entrepreneurship? I’m not sure, and I’m really tempted to have some kids and find out.

(be right back, going to change my OKCupid profile to include “must be interested in n=1 developmental-psych experiments, have access to a rare book library, and speak either Hungarian or Esperanto”)

I mentioned this plan to a friend, who protested that this was cruel and tantamount to child abuse. After all, how can you force someone to spend their entire childhood indoors, studying mind-numbing chess problems day in and day out, instead of enjoying themselves like normal kids?

First of all, this isn’t how the Polgar children (or adults) describe their experiment. From The Guardian:

Starting with his eldest daughter, Susan, Polgár was careful to treat it as a playful activity, turning it into a fantasy of dramatic wins and losses. Whereas Earl and Kultida Woods had coerced perfection from Tiger, the Polgárs encouraged enjoyment, By the time Susan had turned five, she was excited by playing and spent hundreds of hours practising. She was entered into a local competition and treated it as fun, winning 10-0, causing a sensation.

Meanwhile, her younger sisters were intrigued and László allowed them to feel the pieces, seeing them as toys, with no formal tuition until they were five. Interviewed recently, all three girls described playing the game as something that they loved doing – it never felt like a chore. Instead of messing about playing Monopoly, netball or going to the local swimming pool, chess was just what the Polgár family enjoyed…Polgár understood that coercion was less valuable than small children’s need to enjoy fantasy play. Consequently, his daughters all seem to have grown into satiable, well-balanced people rather than success addicts.

But more important – I responded that the Polgars claim to have spent about 48 hours a week practicing chess. I spent seven hours a day in school, so if my teachers assigned two hours of homework a night then we spent about the same amount of time getting educated. Except what the Polgars got out of it was world-champion-level mastery of their favorite subject in the world, nationwide fame, and (by their own accounts) loving every second of it, and what I got was staring out a window all day as my teacher declared that we were going to make a collage about the meaning of Respect.

The Polgar sisters talk about how they loved their education, had a great childhood, thought their parents were always patient with them and never strict and harsh, and don’t regret anything. How many kids who went to public school can say the same?

An article about Laszlo Polgar mentions that he had to fight the Hungarian authorities to be allowed to home school his children. Imagine being so certain of your own home-schooling techniques that you’re afraid taking your kids to the Fasori Gymnasium is going to stunt their intellectual growth. And imagine being right. And imagine my friend thinking that normal American public school might be better than that. It sort of boggles the imagination.

And I guess I shouldn’t be too harsh, because the public school system tries to do the best it can with an impossible set of constraints. But I’m still suspicious. Who else has the motivation to hide that book?

[EDIT: Thanks to readers, I’ve got an Esperanto copy and a person willing to translate it. I’ll let you know as this develops.]

Posted in Uncategorized | Tagged , | 437 Comments

Hungarian Education II: Four Nobel Truths

1. Israel historically has only a moderate number of Nobels per capita

On Friday, I discussed the phenomenon of Hungarian science geniuses, and conjectured it was because of Hungary’s high concentration of Ashkenazi Jews. A commenter pointed out that Israel had an even higher concentration of Ashkenazi Jews, with less impressive results:

By this logic, Israel should have become the hotbed of geniuses. And while it’s true that there are a lot of smart people there, none of the Israeli universities are in the top 10 or maybe even in top 100. And the fraction of Nobel prize winners is not impressive, either.

The objection makes superficial sense. The list of Nobel winners per capita puts Israel at a modest tenth place, after places like Norway and the United Kingdom.

This doesn’t look promising for any Ashkenazi-Jew-based theory.

2. But their more modern numbers look much more impressive

On the other hand, that list counts total Nobels won, ever, and divides them by modern population. That gives an advantage to older countries. Norway’s been collecting Nobels since 1903; Israel wasn’t even founded until 1948. And for the first couple generations the Israelis were pretty busy starting kibbutzim, building infrastructure, fighting off enemies, et cetera. Setting up a good university system capable of churning out Nobels takes time. So Norway and the UK had an unfair head start.

I redid their analysis looking only at Nobels won since the year 2000 (because it was big and round and serves as a signal that I’m trying to avoid optional stopping). My source was this list of Nobel laureates by country, and I deferred to Wikipedia’s judgment about whether or not to count dual citizens, immigrants, et cetera. Here’s the results:

We see that during this period, Israel has by far the highest number of Nobel prizes per capita.

3. This advantage increases if we look only at Ashkenazim

The original theory was about Ashkenazi Jews in particular. Only about a third of Israels are Ashkenazi (the rest are other types of Jews, or Palestinians, or other non-Jewish minorities). If we separate out the Ashkenazim, the graph looks like this.

ISAZ is Israeli Ashkenazi Jews, considered as a separate population. On the one hand, it’s kind of unfair comparing Israel’s most successful population group to other countries taken as a whole. On the other hand, if we were to take other countries’ most successful population groups, those would be Ashkenazi Jews too, so whatever. Since Israeli Ashkenazim get about five times more Nobels per capita than any country, I’m going to consider the “what about Israel?” objection officially refuted.

4. But there’s not a lot of evidence for benefits to concentration, and other factors might be involved

One more graph:

USAZ is US Ashkenazi Jews, who get twice as many Nobels per capita as their Israeli cousins (I’m not sure how seriously to take this; the Israeli data is based on eight Nobel laureates, so there’s a lot of room for sampling issues.)

And although it’s hard for me to get exact numbers, it looks like a lot of Israeli Nobelist (maybe more than half) did their best work abroad, usually in the US.

Israel went from 1948 to 2002 without winning a single science Nobel (it did win in Literature and Peace during that time). Now it’s winning more of them – a lot more, more than any other country per capita – but mostly when its citizens go and study in foreign universities. This seems consistent with an Israeli educational system that’s still struggling to get its act together.

Does this mean that once the educational system gets its act together more fully, the ISAZ Nobel rate will approximately double to match the USAZ Nobel rate? I’m not sure.

Just from this analysis, it doesn’t look like the theory in the last post, where everyone gets benefits from concentrating closer together, is true. Israel has about ten times as many Ashkenazi Jews per capita than the US, but still does worse than they do.

These data don’t challenge the conclusion from the last post that Ashkenazim might have been responsible for Hungary’s sudden crop of great scientists. But they do potentially challenge the implicit conclusion that the education system didn’t matter that much. I’ll have more on that later this week.

Posted in Uncategorized | Tagged | 226 Comments