= 27d567bc2d48d9f40e9ccc20ea370b15)
skeptical_lurker comments on Non-standard politics - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (231)
I put 'other' and wrote 'technocracy', by which I mean rule of experts/replacement of elections with standardised tests. And I don't mean that the country should be run by someone with an IQ of 180 and no social skills, the tests should also measure aspects such as ability to detect lies/lie, emotional control, credence calibration and so forth. The tests and criteria would also be non-uniform, so a foreign minister would require more social skills then the health minister.
I could go on, but suffice to say this should solve all the problems the NRx people identify with democracy, without putting all power in a single point of failure who is chose on the basis of being inbred.
Sounds like Imperial China's civil service exams, with emphasis on different skills. (The civil service exams focused on Chinese classics, which is probably a decent proxy for conscientiousness and intelligence but not so good at measuring social skills.)
There's obviously something to be said for the system, since it lasted for better than a thousand years and weathered several dynastic changes. But plenty has also been written about its drawbacks, and the main ones seem to revolve around regulatory capture: empowering experts to rule also empowers them to decide the meaning of "expert", which we might expect to incrementally make a system less meritocratic and more aristocratic. Goodhart's law is also worth thinking about.
From what little I know, Imperial China was a highly functional civilisation, and as you say, any system that can last a thousand years is impressive.
Of course, far more thought would be needed to set the system up. Perhaps the examination-writing body should be separate from the rest of government? Perhaps it alone should have some form of democratic oversight? Would there be a constitution as well? Overall, I think the best thing is to have fluid intelligence as an essential component of the tests - if the tests focus on Shakespeare and medieval Europe then they can be accused of cultural bias, or if they mainly recruit experts from elite universities then the administration process there wields tremendous political power. But Raven's progressive matrices for instance is completely objective.
Anyway, isn't democracy somewhat aristocratic? The current UK prime minister is the 5th cousin twice removed of the Queen, and politicians tend to come from very posh public schools (one school in particular produced 19 PMs). In the US, the Bush line has been described as aristocratic, and it takes a lot of money to run a presidential campaign.
Um, "cultural bias", i.e., limiting important posts to people who've assimilated into and agree with the culture, was a large part of the point of the Chinese Examination System and a big part of the reason why the system remained so stable.
You sound very confident of that. Is there positive evidence for it, or is it just that it seems plausible?
Nornagest seems to be worried about the system becoming aristocratic, so I somehow doubt that he'd be interested in a more NRx flavoured technocracy. While technocracy could come in many forms, I personally would advocate for a less culture-neutral system - at a bare minimum the test should require fluency in the native language and knowledge of the countries' history.
But I also wouldn't advocate too much cultural bias - after all, we are discussing China in a positive light!
The most important aspect (from the point of view of stability) is making sure people agree with the concept of a test-based meritocratic society and with parts of the culture that support it.
It's equally important that the officials who are chosen that way aren't too awful at their jobs.
I would agree that this is important, with the caveat that a system should be able to evolve to changing circumstances, so there ought to be room for a dissenting voice within the part of the system that controls the systems evolution.
Well, the historical Chinese system wasn't very good at dealing with changing circumstances and dealt with it by discouraging technological innovation.
If you let the system change freely it'll change to a form that causes the meritocratic parts (and even the openness to dissenting voices part) to collapse.
I don't know whether it's possible to combine stability and adaptability. My attempt would be to include an "unquestionable core" to protect meritocracy and the ability to question everything else. But as St. Thomas Aquinas's successors discovered, even that may not work.
What exactly do you think it will change into?
A plausible idea. Essentially the government would have a constitution. Another idea would be that the constitution can be changed, but with differing levels of unanimity needed (so the "unquestionable core" would need a 90% vote to change for example - I'm worried about making anything entirely irrevocable.)
Well to common failure modes are collapse to hereditary aristocracy and "pod people" takeover by fanatical ideologues. The way the later works is that since not "pod people" are willing to hire competent "pod people" but the fanatics will base hiring on ideological fanaticism rather than competence, an ideological faction will gradual take over unless stopped by other forces. For example, look at the current state of academia outside hard STEM fields.
Why exactly history and what do you consider history to be in this case? Dates?
Are you familiar with the adage "those who do not study history are doomed to repeat it"? I am talking about the standard history exams that one might take in school or at university, and when I studied history at school there was a greater emphasis on 'why' rather than 'when'. Its important to know roughly when stuff happened, but only insofar as it helps a general understanding.
And world history in general is important, but your own countries history is especially relevant, so more weight should be attached to it, although certainly not to the exclusion of all else.
You misspelled "Kennedy".
I never said that the same thing didn't apply to Kennedy.
You can easily tune your fluid intelligence test in a way that gives woman an advantage or in a way that gives men an advantage.
No, performance on that test is trainable.
If typically male forms of fluid intelligence are more important for government, then all other things being equal there should be more men in government.
If typically female forms of fluid intelligence are more important for government, then all other things being equal there should be more women in government.
Perhaps you would want a female foreign minister and a male minister of defence?
I was also thinking that most people take the IQ test at 16 or 18, in the same way you have the SATs in the US. In fact, in the US you could just use SAT scores instead. This way, everyone would train for the test.
Who's the "you" you are talking about?
I probably should have written 'one'.
That doesn't solve the issue. Who get's to make that decision?
A lot of mental tests benefit some heuristics over other heuristics. Maybe you find a heuristic that correlates with openness to experience. If your test favors people with high openness to experience you get less conservative people in your government.
Test design is highly political if the results of the test matter.
I guess using the classics is also good for preventing standard drift, since they are set in stone.
But reading the Wikipedia page, the exams seem a lot more subjective than I had imagined, basically entirely based on free-form essays on political problems? It seems this would lead to a ruling class of The Economist columnists, which is a pretty terrifying prospect. :)
I'd love to make a suggestion that your tests include a goal/empathy/altruism/duty test that ensures they're not in the office in order to simply enrich themselves through corruption.
I'm not sure any of those things measure incorruptibility.
Agreed, but I think they'd have some correlation, and I strongly suspect their absense would predict corruptability.
One of the fascinating world-building pieces of Divergent was that the group in charge politically was not Erudite (the clever/curious group), Amity (the kindness/service group), or Candor (the honesty/justice group) but Abnegation (the frugal/self-denying group).
Cheers for the mention. I still haven't worked out if Divergent is meant to be a dystopia or utopia (somewhere in between I think?). Its an interesting world.
I think it's dystopic that they see the virtues as rivalrous instead of cooperative (wouldn't you want someone to have as many virtues as possible, and to 'graduate' from various groups?). The post-apocalypse part is hard to measure; less alienation, but also less trade.
I would suggest, though, that a real teen dystopia is one in which everything is perfect and you are not needed- and so the existence of an obvious defect that you can change (and become important by doing so) seems like a component of a teen utopia.
I think you've got a good point regarding having as many virtues as possible.
On the idea of perfection being dystopic, this reminds me of an argument I sometimes hear along the lines of "evil is good because without evil, good would just be normal", which I don't find very convincing. Still I guess a society and its people should always focus on betterment of themselves, and perfection is probably better thought of as a idealised goal than some place we arrive at.
Its a nice idea, although when making decisions regarding the live and death of many people, an empathic person might simply shut down, so it might be good to include some more dispassionate people who can shut up and multiply.
Unfortunately, I'm not sure there exist tests that can measure these things without being faked. Maybe you could measure empathy by looking at oxytocin levels? But more oxytocin=more empathy is a huge simplification.
Do you have any good ideas for a goal/empathy/altruism/duty test?
I think you're right that the relationship is complex and you probably wouldn't want to optimise just around empathy/altruism. In particular highly empathetic people can sometimes run into problems with cognitive bias around large abstract concepts or numbers. I'm guessing there might be a "sweet spot" for leaders of having enough empathy to want to do the right thing, but not to be overwhelmed by emotion and unable to make difficult decisions.
I'm very interested in possiblities for that sort of test, but it could be tough finding something that can't be gamed. Perhaps some research looking at a range of candidates for (perhaps multivariant) correlations with morally good and effective leadership decisions needs to be done. Actually... surely someone would have done that... though I haven't run into it so far...
A good idea might be to have a mix of cognitive styles so that you can approach a problem from different sides. Of course, you need to be able to decide between these different viewpoints, otherwise you just create arguements.
The first problem is to identify morally good and effective leadership decisions. This isn't easy.
A good idea might be to have a mix of cognitive styles so that you can approach a problem from different sides. Of course, you need to be able to decide between these different viewpoints, otherwise you just create arguements.
The first problem is to identify morally good and effective leadership decisions. This isn't easy.
Yes that's a fairly good point and I don't know any easy way around it either. Looking in the world of business, government, politics etc etc. would be a matter of fairly subjective ideas about moral goodness.
I suppose you could formulate an approach along the lines of experimental psychology, where you could deliberately design experiments with clearcut good/bad group outcomes. So get a bunch of people to be leaders in an experiment where their goal was to minimise their group members (including themselves) getting hit in the head with something unpleasant, build-in some selfish vs unselfish options, and then look at the correlations between leadership behaviours and oxytocin or whatever else you wanted to measure as an input. With a robust range of experiments you could perhaps develop something broadly useful.
I don't think that's the case. EQ seems to make people both empathic and also able to keep a clear head when it comes to tough decisions.
Ideally I also support more testing for everything, but there's a difficulty in keeping more complex tests both accurate and non-corrupted. IQ tests should be fine for some things but probably not everything.
I'm guessing that "here are 20 people, who are each going to tell you a statement. You have five minutes to cross question each one, and then you must decide if they are lying" would be a decent test of social skills for instance.
That test is quite easy to corrupt by telling the 20 people to make it easy for them of the people who get interviewed.
More importantly you test for a very specific skill that's likely trainable. The best people at the skill will be people trained by specific coaches in telling lies in that artificial setting. Those coaches might be payed for by lobbyists.
In real life telling whether people lie when things are at stake for them is a much more useful skill than telling whether someone is lying for whom nothing is at stake.
In most real world contexts where telling whether someone lies is important you analyse the persons motivations and the emotions that body language reveals. That's highly different from telling whether someone whom you give a random card from a deck of cards lies when he says: "This is the Queen of Hearts"
You could make there be something at stake by having everyone telling lies also be part of the group taking the test.
Unfortunately(?), this would mean that people would train at telling convincing lies as well training at detecting lies.
I second the question mark.
This kinda sounds like 'trials are easy to corrupt by telling jurors how to vote'. If this specific test were implemented then it would be possible to, for instance, choose the 20 people by lottery. Maybe look into how people prevent jury tampering?
I don't think the primary goal is to have a system 100% free of corruption. That simply isn't realistic.
I was more thinking of having people lie about statements about their lives, perhaps ones with emotional significance, to make the test more realistic.
Jurors are supposed to be able to use their subjective assessment of the situation. If the juror thinks that the person shouldn't be judged guilty, then the person should get an advantage.
Furthermore getting a single jury case wrong isn't as bad as selecting the wrong president. It's more important to have a system for selecting presidents that immune to tampering.
There's still no motivation to lie in that example besides: "The experimenter assigned me the role of a liar."
Ok, so I think that what you are saying is that if the person being tested is already a public figure (like a president up for periodic re-testing), then the 'liars' will have pre-existing opinions destroying the objectivity of the test. This is a good point - perhaps this specific test can only be used on new applicants.
When it comes to selecting a president, extra care must be taken. One possibility is that testing is used to select people at the senator level, who then vote one of their number into the presidency. This means that many tests would have to be tampered with to subvert the system as a whole.
You could give the liars a financial incentive. You could play candidates off against each other. You could get them to literally play the game 'diplomacy'.
I think it's very likely that there are special skills involved in the game diplomacy.
There the notion that American politics often resembles poker while the Chinese rather play go. I don't know what heuristics come with the game diplomacy but those heuristics could also matter for politics.
The whole situation also raises a bunch of stress in applicants which can cloud the body language.
Are you aware of any company for which hiring people with high social skills is important who let's their applicants play diplomacy?
I think there is a reasonable case that go teaches certain useful skills beyond 'just' providing generic brain excersize. You have to know which groups to fight for and which to abandon, you have to prioritise, you have to avoid becoming fixated on any one part of the board. 'Play urgent moves before big moves' is good life advice.
Diplomacy might train people to avoid being stabbed in the back, or it might train them to stab other people. You could even invent your own, positive-sum game if this seems like a potential problem.
Politicians are going to need to make decisions under stress, and deal with stressed people.
Since what I am proposing is similar to a job application process, looking at the hiring process for high-paid corporate roles could be a good starting place for anyone who was actually trying to implement this in real life, as opposed to my attempt here to paint a rough picture of what the process might vaguely look like.
I do know that some financial companies have 'the theory of poker' as required reading, and a quick search turned up this recent idea of using custom video games but I think in general companies use interviews more.
Of course, using interviews to select politicians simply allows the government to form an aristocracy. Companies are at least accountable to their stockholders. The idea of turning the government into a company and giving the people shares was, I believe, an idea of Moldburgs'. I find it more interesting than reverting to monarchy, but it has its downsides, especially that, given the distrust of the financial system, I cannot see it having to popular support to get started in the first place.
'Play urgent moves before big moves' is not so much what I'm talking about. Go strategy suggests that attacking weak positions directly is a bad idea. In Go power doesn't get used to bluff. You continue to build power and if you are strong enough your opponent has to sacrifice a few stones because it's not worth to defend them.
China's idea with Taiwan isn't to take it in a bloody war. It's idea is to get enough power that Taiwan has no other choice than to come back. Chinese foreign policy is different than US foreign policy.
Poker has no notion of aji keshi and a lot of people in the west don't use a concept like aji keshi in strategic conflicts. I'm not even aware of a good word in the Oxford dictionary for aji keshi.
I don't know about strategy behind Diplomacy but given that I have never read a book about Diplomacy strategy that explains why I don't know. If you have never learned Go then given the knowledge of Go rules you wouldn't come up with the concept of aji keshi yourself and see that it's important in Go.
If I have a high stakes negotiation then I can usually safely assume that the other person is stressed because he cares about the outcome of the negotiation. If he is on the other hand stressed because his wife send him an SMS right before the negotiation that she wants to divorce and he spends all the time thinking about the SMS instead of focusing on the negotiation then he becomes hard to read. Especially if you don't know about the SMS that person get's very hard to model.
If I do hypnosis and say a wrong word then the tension in the person I'm hypnotising rises. I perceive that change in body tonus and can change course. That helpful as long as the person doesn't get suddenly tense for reasons that have nothing to do with my interaction with him.
As long as I'm having a decent mental model of the other person and perceive body language I can sometimes do well as far as mind reading goes. On the other hand I lose that if there are stress factors I can't decently model.
"Popular support" might not be needed. As multinational corporation get stronger and nation states get weaker we might get a world where a corporation get's stronger than a state. It's possibly that a corporation just overtakes a powerless African state.