- With lots of eyes, the median position is most likely to be right. Adaptive markets FTW.
- Smart folks generate, and promote, and argue for alternative hypotheses.
- Different smart folks generate conflicting hypotheses.
- Most of the hypotheses generated (and adopted) by smart folks are wrong.
- Occasionally, relatively independent of the intelligence of the smart person, a hypothesis pans out.
- The lucky, unusual, correct hypotheses are the foundation for all of the difference between our lifestyle in America today, and that of the dirt-eating Mesopotamian peasant 5000 years ago.
- IF smart people weren't generating dozens of wrong hypotheses all the damn time, we'd be a lot worse off as a society.
- However, on average, the average smart person should not expect that anywhere near half of their contrarian positions are correct.
- Really, this is a particular case of the generalized competition-benefit. Trick lots of folks into playing positional (status) games, and get 2nd order benefits to everyone else, while the competitors waste huge amounts of energy mostly losing. And then we celebrate the winners, to get more folks to do the same thing.
Tuesday, April 10, 2012
On Smarts and Error
Subscribe to:
Post Comments (Atom)
Posts
36 comments:
You and your war against the peasantry! We do not eat dirt. We mix it into our smoothies. It helps guard against asthma.
The French word for peasant is 'paysan' and it is used by our neighbors to describe themselves. It isn't as negative as 'peasant' but something more like 'simple farm folk'.
Wobbly,
Read the comment out loud to my wife
"IF smart people weren't generating dozens of wrong hypotheses all the damn time, we'd be a lot worse off as a society."
And yet, we think it's smart to channel all of our resources into the conspicious consumption of a few lucky smart people rather then giving the great mass the resources needed to promote more ideas.
Going with the average belief is a good idea in nonpolitical questions, but unfortunately most controversial questions are political to some degree. In many cases people have an incentive to profess the fashionable belief to raise their status and affiliate with desired allies. In those cases, stated belief is likely to become *systematically* biased, and going by the average will skew you away from the truth.
Of note: smart people are not just better at rationalization, they're also better at navigating complex social dynamics...such as figuring out which opinions will raise their social status and which ones will lower them.
"Of note: smart people are not just better at rationalization, they're also better at navigating complex social dynamics...such as figuring out which opinions will raise their social status and which ones will lower them."
I think the second sigma is the maximized point for this. Past that I find there is far more intrinsic value put on truth (can't believe your own rationalizations so easily).
Also, since very high sigma have difficulty communicating with normals they are less able (and thus willing) to play status games.
Anon1,
I'm not quite sure what you said.
I think you just made a gratuitous attack on the notion of private property: When folks trade with you, because they want what you trade more than what they currently have...that this is basically fair.
AC,
Well said. There are real limits to the social epistemology that I'm pushing...most notably when there are low costs to being wrong, and when there are alternative reward systems that reward things other than correctness. Hence my adaptive markets plug, which handles most of those issues tolerably. Regardless, well said, you.
AC & Anon,
I think Anon has the stronger of the arguments pointing out the max social dynamics @ ~2Sigma.
On the other hand...I find the line that high sigma folks are less able to play status games laughable. They play status games just like anyone else, but usually only in their cliques, and with more "I'm smarter" status. I don't see that directly impacting anon's argument, but it might.
However, Anon's line that smart folks can't believe their own rationalizations so easily is full-on nonsense.
Aretae,
What I'm saying is that if progress comes from smart people throwing ideas at a wall and a tiny minority sticking, the way to maximize progress is to get the most ideas being thrown at the wall. How do you do that?
1) You need incentive to throw things.
2) You need to give people the means with which to throw ideas at the wall (time, education, resources).
3) You need power distributed equally enough that people can throw their ideas at the wall without it getting knocked down by some 800 pound gorilla.
Further, lets assume that most markets in ideas are winner take all, as they are now.
Well, with no intervention you've got lots of incentive (the winner gets everything), but then every other party doesn't have the resources to come up with ideas (#2) and if they do the guy with all the resources knocks them down (#3). We could share equally with everyone, but then there is no incentive (#1).
Some mix of laws and incentives is required to meet all 3.
Aretae,
Well, you've probably talked to a lot more +3/4/5 SD IQ people then me. I mostly interacted with some of them going to school, and a couple when I was trading on WS. Other then when I was dating the Harvard girl, it has been a long time since I've interacted with a genuine genuis.
Perhaps I'm giving +3SD people more credit because they are less likely to buy the PC line, and seemed more inclined to care about how their actions affected others.
I've met some +3SD people who, like me, turned down wall street riches over moral concerns. I've never met a +2SD person who turned down a chance to get ahead. Or who didn't buy into some variant of the cathedral orthodoxy hook, line, and sinker without question. Now that I'm surrounded by +2SDs all the time maybe I'm just idealizing the past a bit.
Anonymous,
The second sigma---ie. from around 115 to around 130, seems to be about the peak of indoctrination.
I'll also confess that though I've met lots of really smart people, I've never met a neurotypical smarter than 3 sigmas. I'm sure they exist---but some of the 3 sigma 'neurotypicals' I've met actually aren't, and are just running really excellent emulation (Bill Clinton fits here I'm fairly certain).
I think the real distinction is that the 2nd sigma thinks they're normal, just better at it---and they aren't.
Those smarter than that recognize the fact that they're pretty alien to most of the rest of society and make the projection error less often.
I was going to ask about the common interpretation of the word 'neurotypical' since I have started to see it in the last couple of days. It seems to be defined by usage such that aneurotypical = weird geek.
Am I right? Are there nuances missing?
Wobbly,
Even "weird geek" can be neurotypical. Geeks are just one of many branding options people take. Do you think the teenagers shopping at Hot Topic are any less neurotypical then the ones at Abecrombie? Geeks fit in, they just fit into the geek tribe.
In browsing for Mad Men on demand last night I pulled up an episode of "New Girl". The character, played by professional geek archtype Zooey Daschel, is a "weird geek." She's also like every other "weird geek" 29 year old manic pixie dream girl teacher girl out there. Her "quirks" are socially approved. The show can get millions of viewers because there are millions of viewers that relate to the character. That's pretty neurotypical.
The best way I've ever seen it broken out is into dorks, geeks, and nerds. Dorks are mutant retards at the bottom on the pile. Geeks are the 95% inbetween that are neurotypical but "quirky" in ways that really aren't all that quirky when you think about it. Nerds are usually really smart people who are non-neurotypical. Some run really good emulation though, as Jehu puts it.
Wobbly,
Anon says it pretty well. I think that history would clarify as well:
Neurotypical and non-neurotypical ar terms that Autism-spectrum folks began to use to clarify that their brains simply work differently. Some of the well-above 2 sigma folks who appear to have similar systematizing tendencies (Simon Baron Cohen's Autism work) have also started to adopt the label as well.
Geek, dweeb or spaz?
From SNL
I wonder whether I am the most neurotypical smart person I know, or my emulation is so good it fools even me.
I feel the alienation in the idea landscape - it is very rare that I find people that I can talk to in the way that I want to. But I have little problem talking with people about the things that they are interested in, in the manner in which they are accustomed. Though I often find myself modeling their speech, to predict what they're going to say next. It's a weird sort of constant deja vu when people say exactly what you expect.
Sometimes I wish I had an asberger switch that I could turn on to allow me to focus better. If you've ever read A Deepness in the Sky (and if not, you should) the concept of Focus as used by the bad guys is at the same time frightening, nauseating and appealing.
I generally use neurotypical to mean 'thinks like 95% of humanity'. Such people appear normal to most of the population, even if they are smart or very smart. They process sunk costs like regular people, they have regular people risk aversion, and they have diplomatic models (that is, what makes them like a person more or less) like most people.
The set of non-neurotypicals includes obviously Aspies/Autists but also most any 'really odd duck' as well.
It's not like a switch. It's not you are or aren't neurotypical. Everythings a spectrum with marginal differences. Someone at the 95.1% of neurotypicalness is going to seem more normal then someone at the 99.9%.
As pefidy put it, I can interact with neurotypical people in a nuerotypical way, I just don't get much out of it. I can do it, and sometimes, in the right circumstances (especially if I drink), I can enjoy it. But generally speaking I just don't get much out of it. Neurotypicals really seem to enjoy what they do. I just...don't.
Re: NNTs
It's also the source of the extremely common superbright feeling: "I'm an alien." When your thought processes don't map in any credible fashion onto what other people say (read: appear to be thinking)...at some point you start to wonder if there's a category difference between yourself and people...and your capability to empathize with others who are deeply *not like you* in the ways you think are important starts to drop off.
I think I've almost completely avoided this mistake, mostly based on watching other folks make the mistake, and observing the costs of said opinion in me...but I did it via a cognitive hack at age 15, not via normal human experience of "like me".
Also probably goes a long way towards explaining my universalism (and that of other optimist high-sigmas like Caplan). We don't naturally feel similar to anyone... but have actively altered our priorities to emphasize our commonality with all people... But the commonality we have focused on is human commonality, not commonality with our race/neighbor/co-nationalist/co-religionist. We are connecting on basic humanity, because we have a hard time connecting on any other aspect. And if basic humanity is the connection...the particularist stuff is all violence against humans.
Aretae,
When I read Caplan* I get:
He feels superior to and dislikes all people equally. I mean, that's universalist, but its narcisistic universalist.
I also think there is a certain leapfroging. Its very common to find people that don't want to practice empathy to those around them (which would require daily real acts of empathy) and justify it with a more universal empathy (its ok because I'm saving the world). So Caplan doesn't have to feel bad about paying his garder slave wages, he's actually improving his live dramatically.
In fairness, the reverse bias is true too. The mob boss that kills people for a living but is a great family man.
*Please try not to project all this onto yourself, I know little about you and haven't read as much of your stuff.
Interesting comments here and probably enough to get the idea although there are a few references I miss - Mad Men I've heard of but Hot Topic? Zooey Deschanel? Never seen one of her flicks. At least I know what a MPDG is, though probably from something Aretae or Isegoria linked to way back.
There’s a good buddy of mine who is an Aspie and there are clear differences in the way he worked in meetings compared to other folk. It would have been funny if it wasn’t taken as so confrontational. I have known and worked a very large amount of very smart people and a lot of them do seem to put on an act. How much is genetic and how much is from growing up and working in a sheltered workshop where idiosyncrasies are celebrated?
"Also probably goes a long way towards explaining my universalism (and that of other optimist high-sigmas like Caplan). We don't naturally feel similar to anyone... but have actively altered our priorities to emphasize our commonality with all people..."
That's the story I'm going to go with from now on. It explains proggie behaviour and my behaviour. And it makes sense when I follow it backwards.
Now I just have to work out whether, considering how much damage proggies do and considering that the evo default is the opposite, if it might be worth actively avoiding it...
One time I was practising the predicting of conversation, and I got a sort of pre-echo. I wasn't a participant, just listening. Anyone else ever see that? The timing was about right for getting the audio cortex to leak directly into consciousness, as if by resonance on the fire-together, wire-together principle.
Anon,
Re: what you meant
I'm largely inclined to see the same problems as you do, with a further note that (1) doesn't seem to be *too much* of a problem ever, and (2) doesn't seem to be much of a problem in the rich modern world. (3) is the key problem from my PoV. And 80+% of #3 can be solved by abolishing patents.
Iff strictly necessary, I would move towards a prizes, not patents model of innovation...perhaps a 1% income tax, distributed among innovators according to some formulaic algorithm analyzing inventions over the last 10-40 years which computes the invention's usefulness, and partitions out $ from the 1% based on said usefulness.
Hypothesis Testing
Define Hypothesis, what is Hypothesis? Define Hypothesis Testing, null Hypothesis,
http://www.infoaw.com/article.php?articleId=952
Anon, by drinking you're making yourself artificially stupid. That's why that works. Ever try to explain something in your normal way while drunk? Can't maintain the thread because its too involved for your impaired brain? That happens to me every Tuesday about 8:00 after my weekly gin and tonic fueled seminar with my friend Chris.
Alrenous, I think you and I are talking about more or less the same thing.
The way I feel when it happens is closest to deja vu, which is why I mentioned that. But what actually happens is, I think, closer to what you're describing.
I do project a bit...I think Caplan has a reasonable amount in common with me. Could be over-identification with folks who have similar thought-patterns.
I don't think Caplan dislikes anyone particularly... indeed, he has explicitly said that he considers the rather standard disliking of people by smart folks an error (maybe even a moral defect).
Also...I believe that Caplan has thinks that other people have worse opinions...but there's a long distance between "My opinion is better" and "I'm better". I know folks in both categories...and Bryan's optimism fits much better into the former category.
As to paying the Gardener wages below what you want them to earn...the choice is not what you make it. You can pay someone wages that they'd be happy with for the work...or you can *think* about paying them the wages you'd like to pay them...and not hire them at all, because it's way not worth it for you. Because those are the two actual choices...pay them low, or don't pay (trade with) them at all...it's hard to see how leaving the Gardener broke is better for him.
Aretae,
Two separate replies:
The first
1 and 3 are in conflict. I happen to agree on patents, but at least there is a tradeoff and a debate.
2 is only true for some first worlders, and as I've harped on many times only for the healthy in America (depending on your risk tolerance and the level of sickness).
The 2nd
Isn't Caplan perfectly obvious in his opinion:
"I find my society unacceptable. It is dreary, insipid, ugly, boring, wrong, and wicked."
I mean we don't need to guess. Keep in mind his lead in wasn't about the politics or anything. It was about hating NASCAR and all the rest. He really does hate the proles, except when they pump his gas.
My guess is if Caplan had to pay more for his gardener, he'd still get a gardener, so long as we are talking within reason. Caplan would consume less of the pie, the gardener would consume more of the pie, not the end of the world. There is likely a large surplus between what he pays the gardener and the utility of the surplus. I'd prefer the gardener capture that surplus rather then Caplan.
Personally, while I think there is a link between productivity and wages, I don't think its as iron clad as people think. Wages are the result of relative bargaining power. Power sets the wage. I never got a wage raise because I was more productive, I got it because I was in a good negotiating position. Being more productive is only one way of getting there.
I certainly have known many people that aren't very productive, but they are in positions of power and exploit them. You can call them rents if it makes you feel better, but the economy is loaded to the gill with rents. All the good jobs involve rents. All the big industries involve rents. When I see those in a position to exploit rents whittle those not in the same position down to the bone so more of their rent is left over for their own consumption I cry fowl.
We live in a rent based economic system. Never forget that. When people tell me they are richer because they are more productive I usually check that list twice.
Anon,
Re: Caplan.
How long have you been reading Caplan?
Me: About 10 years...and we've been running in parallel intellectual circles for 15+. He doesn't hate the proles. He does oppose the people who hate the proles, though. Yours is a narrow reading of 1 comment he's made over the last several years.
This one (against high-IQ misanthropy) is far more indicative of his whole position than the example you use.
Anon,
Re: innovation
#2 is true of damn near anyone who cares*** in the first world.
Of course...it is roughly true that you need a combination of 4+ (maybe 5+) Sigma total of IQ and willingness to throw 90 hour weeks at a problem ( some mix of conscientiousness, stubbornness, patience, give-a-shit-ness) in order to check if your idea sticks.
Except for the rather rare case like yourself...the problem is usually a dearth of willingness to throw 90 hour weeks at a problem for a few years. That's actually the #2 that you need...and at a well-above 90% probability (you excepted)...it's personality issues, not health, that prevents folks from doing this.
2. Gardeners. Your analysis MAY be correct for Caplan. With an awareness of marginal thinking...my line is clearly true across people. It is better to have a society wherein gardeners are cheap, and many get hired, than one where gardeners are expensive, and a lot fewer of them get hired ... and those are the only two choices.
Productivity:Wages is a cross-country comparison, and a longer-term line. At the individual level...it's usually a question of replaceability + willingness to negotiate.
Personally on productivity...I've been independent for most of the last 20 years...and folks have paid me entirely on negotiation & scarcity. There aren't many folks can do what I do. And I get paid for it. On the other hand, folks in any large organization are mostly paid based on bureaucratic status, and almost not at all by merit.
Aretae,
I want you to think this productivity thing through. Here is a very simple example:
When graduated college with other high sigma IQ people, wall street came calling. Everyone knows that they are evil, but they roll up a dump truck full of money at 22 year olds and say, "here's a live path to the 1%." It's a hard choice. Do something productive but risky (the payoffs for start ups in a winner take all system). Or take the sure thing you know is wrong.
I noticed a pattern in how people made their choices. Those with rich parents where more likely, all other things held equal, to try and follow their dream. They knew, implicitly, that if they failed their parents would catch them. The poorer kids with more obligations and financial risk, like myself, choose wall street.
Now, your opinion on wall street productivity may differ from mine, but my impression was that it was one giant massively counter productive enterprise. You could slash compensation by 80% and people would still issue stock and bonds and do mergers and such. So just being smart and putting 90+ hours a week into something doesn't mean your being productive.
What I think your missing here is that its much easier for smart people to use their smarts to extract rents then create anything. That's just the way the world works. The better their position, the less life risk they exposed to, the more they will prefer following their hearts (usually productive) over taking the more sure (usually rents) path to the good life.
That's what I don't think your getting here. When you hollow out the middle class. When you make the economy winner take all. When you make the mating market winner take all. People are under immense pressure to turn to the dark side. Because second place is last place in such a world.
They'll put in their 90 hours, trying to pull one over on someone else.
Aretae,
Caplan can post that, but what are his actions. He wants to build a world in which proles live on subsistence wages and maximize his utility. That's what every single one of his policy proposals is about. And he doesn't care, because those proposals don't effect him, because he lives in a bubble.
For me, I've found that there is often a vast difference between what people value something at and what they pay. I believe who captures that surplus comes down to power. And what I see being advocated are rules designed to maximize one groups power over another.
Anon,
You project your opinions about what kind of world would result from Caplan's preferred policies onto Caplan's actions.
If you read Caplan closely, he can back all of his opinions of how the world should be back to how it would be (according to his knowledge of economics, which is substantial) better than the status quo for the average person.
His motivations are general positive regard towards people. Not seeing that (and not recognizing that he may well be smarter on the topic than you are) is an epistemology problem.
1. Agree...90+ doesn't = productivity.
2. However, 90+ is the prereq for inovation-testing/entrepreneurialism.
3. I think Wall street does incredibly important jobs (moving money from lower-productivity uses to higher-productivity uses)...and some jobs that are substantially less so.
4. I agree that rent-seeking is the problem. That's a major component of my libertarianism. So long as government has enough power to make economic decisions, it will be cheaper for companies to buy the legislators than it will to do "honest" work.
Aretae,
Last reply, trying to pair down topics.
Rent seeking can happen in the "free market". Because textbook free markets just don't exist. I think that is the difference between you and me. You think rent seeking is strictly limited to government. Get rid of it and it goes away. I'm not convinced. Its just so so so easy for high IQ people or connected people to take advantage of the less fortunate. It's all I've seen all day long my entire career. Sometimes the government was there, but a lot of the time it wasn't.
Monkeybrains are not rational actors. And sometimes markets in big societies are, due to logistics, so loaded with principle/agent and asymmetric information problems that they hardly resemble "free markets". Some are better equipped to take advantage of this fact then others, even if there is no government.
Nor do I think you can ever be rid of government (government will always be a Nash equilibrium). If you don't have the ability to win* in the free market (and most people don't, its winner take all) then your always going to have an incentive to band together with others and form a government.
*Herein I define "win" relatively. Any man today is wealthier then Ceaser, but Ceaser was happier then any man today. Because we are social animals, and our happiness comes from having more then others. It means more love, more respect, more pussy, etc. Absolute increases are commendable, but absolute increases aren't enough to organize a society around. It doesn't work that way, the monkeybrain doesn't care. So even if you came up with the perfect libertarian government that maximized everyones absolute level of wealth...people would still tear it down. They would tear it down so long as they had incentive and means, which inevitably would occur. And the incentive would be their own social power.
"I'm not convinced. Its just so so so easy for high IQ people or connected people to take advantage of the less fortunate."
True.
As a market anarchist, I don't see a problem, I see an opportunity to sell scam prevention. Pit me against the rent seekers, but on the side of the potential victims.
No, I'm not skimming the rent, I'm providing security. Might as well say that prisons charge rent for keeping muggers off your lawn.
"Nor do I think you can ever be rid of government (government will always be a Nash Equilibrium)."
"I assume freedom==anarchy is impossible. Therefore, anarchy is impossible."
Well...yes. Yes, that definitely follows.
Alrenous,
I think you overestimate the effectiveness of the scam protection market.
Example, prisons keep muggers off your lawn. They also lobby congress to make drugs illegal so that they can make money arresting and storing drug users. You invent one institution to prevent a rent and it becomes its own rent seeker. Rent seeking is human nature. Its an equilibrium result.
Post a Comment