T̶o̶ ̶s̶u̶m̶m̶a̶r̶i̶z̶e̶ ̶t̶h̶e̶ ̶a̶r̶g̶u̶m̶e̶n̶t̶ ̶m̶a̶d̶e̶ ̶a̶g̶a̶i̶n̶s̶t̶ ̶Y̶u̶d̶k̶o̶w̶s̶k̶y̶'̶s̶ ̶r̶e̶a̶s̶o̶n̶i̶n̶g̶:̶ (see my comment below on why the strikethrough)
Here's a summary of the original counter arguments that commenters have brought against Yudkowsky :
1 - Hillary is not necessary a better B player than Trump
2 - existing level B players might deluding themselves that they are making a good job at maintaining the status quo
3 - the local optima might be surrounded by a valley of worse equilibrium, but only if you care about the wellbeing of the whole world equally: if you care only for America's interest, then other equilibria might be more beneficial
4 - Trump was actually testing the waters with his moves, when elected he is going to revert to a saner, albeit different, policy
5 - he is weighting personal experience regarding level B too much, possibly such a level don't exists or its importance isn't too high
6 - Trump's move wasn't bad: Yudkwosky didn't take into account the support of other intellectuals in the same sphere
7 - the letter of those condemning Trump weren't motivated by real preoccupation with Trump declarations but by political enmity
I'll group the answers thusly:
Thank you to everybody.
This was a very productive and high quality discussion, to me a strong proof that LessWrong is far from dead.
Upvoted. Thank you for hosting.
6 - Trump's move wasn't bad: Yudkwosky didn't take into account the support of other intellectuals in the same sphere
Both user:hg00 and I argued that Eliezer stopped searching for expert opinions in a motivated way, but I concluded that relying on expert opinion, which ultimately appears to indicate that Trump will probably have more negative effects on our foreign policy than Clinton, was correct anyway. The OP specifies that the purpose of the discussion is to evaluate methodology, and remains silent on the evaluation ...
Sorry for the slightly clickbait-y title.
Some commenters have expressed, in the last open thread, their disappointment that figureheads from or near the rationality sphere seemed to have lost their cool when it came to this US election: when they were supposed to be calm and level-headed, they instead campaigned as if Trump was going to be the Basilisk incarnated.
I've not followed many commenters, mainly Scott Alexander and Eliezer Yudkowsky, and they both endorsed Clinton. I'll try to explain what were their arguments, briefly but as faithfully as possible. I'd like to know if you consider them mindkilled and why.
Please notice: I would like this to be a comment on methodology, about if their arguments were sound given what they knew and believed. I most definitely do not want this to decay in a lamentation about the results, or insults to the obviously stupid side, etc.
Yudkowsky made two arguments against Trump: level B incompetence and high variance. Since the second is also more or less the same as Scott's, I'll just go with those.
Level B incompetence
Eliezer attended a pretty serious and wide diplomatic simulation game, that made him appreciate how difficult is to just maintain a global equilibrium between countries and avoid nuclear annihilation. He says that there are three level in politics:
- level 0, where everything that the media report and the politicians say is taken at face value: every drama is true, every problem is important and every cry of outrage deserves consideration;
- level A, where you understand that politics is as much about theatre and emotions as it is about policies: at this level players operate like in pro-wrestling, creating drama and conflict to steer the more gullible viewers towards the preferred direction; at this level cinicism is high and almost every conflict is a farce and probably staged.
But the bucket doesn't stop here. As the diplomacy simulation taught him, there's also:
- level B, where everything becomes serious and important again. At this level, people work very hard at maintaining the status quo (where outside you have mankind extinction), diplomatic relations and subtle international equilibria shield the world from much worse outcomes. Faux pas at this level in the past had resulted in wars, genocides and general widespread badness.
In August fifty Republican security advisors signed a letter condemning Trump for his position on foreign policy: these are, Yudkowsky warned us, exactly those level B player, and they are saying us that Trump is an ill advised choice.
Trump might be a fantastic level A player, but he is an incompetent level B player, and this might very well turn to disaster.
High variance
The second argument is a more general version of the first: if you look at a normal distribution, it's easy to mistake only two possibilities: you either can do worst than the average, or better. But in a three dimensional world, things are much more complicated. Status quo is fragile (see the first argument), surrounded not by an equal amount of things being good or being bad. Most substantial variations from the equilibrium are disasters, and if you put a high-variance candidate, someone whose main point is to subvert the status quo, in charge, then with overwhelming probability you're headed off to a cliff.
People who voted for Trump are unrealistically optimists, thinking that civilization is robust, the current state is bad and variations can definitely help with getting away from a state of bad equilibrium.