@MathWizard's banner p

MathWizard

formerly hh26

0 followers   follows 0 users  
joined 2022 September 04 21:33:01 UTC 

				
User ID: 164


				
				
				

			

		

	














	


		


			



	

		

			

			

				

					

						

							
						

						

							
							

								
MathWizard


								

								

								
								
								

								
							
					
 
				


				
formerly hh26

				
				
				

				

					0 followers  

					follows 0 users  

					joined 2022 September 04 21:33:01 UTC
				


				
					

					
No bio...

				

				

				

				

				

					

					

					
					
					

					
User ID: 164


					
					
					
				

			


					

						
					



		

	

 











			
			

			
			

			




	


		

			
‎

			

				
				

					Hour
					Day
					Week
					Month
					Year
					
				

			


			
‎

			

	
	

	 Hot Bump Top Bottom New Controversial Comments
	



		

	






	

		
		

		
		
	

	

  







		

			
			
				Small-Scale Question Sunday for September 18, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					


I do not consider repeated game strategy to be "out of game". It's a basic element of game theory - ever heard of repeated prisoner's dilemma?




In which case the "Game" your are trying to optimize for is the sum payoff over the entire repeated scenario.  That is, you have one main game, which is composed of many subgames, and acting rationally within the larger game may involve local "irrationalities" in the subgames which are only rational within the larger structure.  Importantly, this is explicitly declared in the game formalism.  Individual board games do not mention each other, so unless you're at like a board game tournament or doing a best 2 out of 3 or something, each game is being considered independently.  Maybe you as a human being want to maximize the number of board games you want to win or something, but actually you also want to optimize for things like money and friendships and comfort and happiness and eventually we've gone full meta.  My claim isn't that it's impossible to treat board games as repeated games or that you won't improve your winrate, my claim is that it's inappropriate and unfun.  It's effectively a defection in the board game playing experience, something which increases your own enjoyment (assuming you like winning) at the expense of everyone else, and if everyone does it then everyone ends up having less fun.




Kingmaking for game theory purposes, especially if warned beforehand, is valid strategy. Introducing strategic spite into the game makes the table rethink how they build alliances and gives players more agency.




Conditional on you being able to keep the spite entirely within the game, and credibly signal that to other players so they're not worried about making you upset in real life, I would again consider this to be a defection in the board game playing experience.  It will make you more likely to win, and make the game less fun for everyone else because now you're restricting which actions they can do.  If you unilaterally declare an ultimatum "nobody can do any harmful actions against me or I will sacrifice all chances of winning to destroy you" then you'll have a massive advantage as no player wants to incur your wrath (unless they're so far ahead they can afford it, or so far behind they are going to lose anyway and want to reverse-kingmaker you).  But if everyone player does this then you have a big mutually assured destruction scenario and, unless the game was specifically designed around that scenario, is likely to be less fun than playing the way the game was intended.


Note that spite being a defection, a form of unsportsmanlike conduct, does not mean you should literally never do it.  I would consider it appropriate in a meta tit-for-tat scenario, where you threaten players who behave unsportsmanly against you with unsportsmanly spite.  If a player seems to be picking on you unfairly and spitefully, or doing some other action that is legal within the rules but the entire group agrees is bad behavior, then you can spite them back to punish their behavior.  But in general the best outcome is one in which everyone cooperates, which means voluntarily forgoing a small set of behaviors that are technically legal but unfun, which varies from game to game but generally includes most meta concerns and kingmaking.  You should generally seek to increase your chances of winning, but not goodheart it at the expense of having fun.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 3, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Sure, but it's a relatively simple computer program.  And you can write those to run perfectly the first time if you are very careful, and if the stakes are high enough to incentivize you to double check your work before submitting.


Importantly, the ability to do this is a skill which can be learned, and is important to actually use when making predictions.  When I see




90% All mask and vaccine mandates in all Western countries will lift by September.




a red flag goes up in my mind.  Because "all" is an extremely ambitious condition, and 90% seems way too high for that.  And part of the point of being rational (or rational-adjacent) is to recognize and avoid the exaggeration and hyperbole that everyone else uses in common speech.  You might casually say "all of mandates will be gone by September" and, when someone calls you out and questions that as being unrealistic, and asks for a concrete prediction, you should think about it more deeply and walk back the exaggeration.  "Well, not literally all, they'll probably keep some for healthcare workers, and maybe one or two nations will keep most of them, but I predict at least 8 out of these 10 specific nations will lift mandates for 90% of the population" or something like that.  The fact that this person didn't walk back their bold and unrealistic claim when making a bet is an actual mistake that deserves a loss, not a technicality.  The term "all" didn't set off a red flag in their mind, and it should have.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 10, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I'm mostly agnostic on HBD (though I lean pro) precisely because I don't believe the world would look all that differently if it went one way or another.  My  is that the majority of issues in poor minority groups are caused by culture: a lack of respect for education, marriage, rule of law, and unselfish cooperation with each other, and that these cultural elements are self-perpetuating and economically crippling.  While innate intelligence does play some role in influencing whether a person will abandon or change these cultural elements, it's a minor role.  People with a genetic predisposition for high IQ but a bad culture frequently end up in bad outcomes because they fail to learn or care about learning and never rise to their true potential.  Similarly, low IQ people with a good culture often become productive workers and good people and beneficial to their community because they work hard and care about people.  IQ plays some role, but culture plays a much larger role.


Society is filled with selfish intelligent people and kind unintelligent people of all races.  But they tend to come in clusters, as culture perpetuates these traits separately from genetics (though still tending to run in families), so you see disproportionate amounts of selfishness and other negative cultural traits among certain races.  Heritabile =/= Genetic, and the distinction is important because culture can change, while genes can't.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 10, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
You always have to be careful about controlling for confounders, but there's enough evidence in the same direction that I generally buy it.  HBD is probably true, but my argument is that its effect is significantly smaller than the effect from culture, so it's not an important priority for addressing or using to explain gaps.  It's not as simple as reasoning "Median househould income is $77k for white people and $46k for black people, but white people are smarter so everything is fine".  If HBD is false then with equal cultures, and absent racism, the median income for black people would also be $77k.  If HBD is true, then with equal cultures the median income with equal cultures might be $72k or something, something between $46k and $77k and closer to the latter than the former.  The gap is caused by multiple factors, and there is significant progress that can be made, and most but not all of the gap could theoretically be closed.  If HBD is true, then it will eventually be important to acknowledge as true so that someday if we reach the equilibrium we don't keep endlessly looking for racists and/or cultural issues, because the gap can't ever be closed completely.  But at the moment there's so much other stuff going on that it's only a small piece of the pie.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 10, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Really?  Because the stereotypes I'm familiar with involve Jewish mothers hounding their kids to excel and overachieve and never being satisfied with mediocre or average results (maybe I'm mixing this one up with Asian stereotypes, but I think it's true here to a lesser extent).  Also, intellectual arguments and thought experiments about Jewish law, and respect for cleverness.  And rather than being purely selfish, the stereotype I usually hear is that Jews are selfish in dealings with gentiles, but friendly with other Jews, preferentially hiring each other due to a sense of shared culture and nepotism.  Which is probably even more beneficial for a subculture with enough power than general unselfishness would be.  I don't know how truthful these stereotypes are, but to the extent they are true I would predict that they would lead to above average success.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 10, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
It's a repeated prisoner's dilemma: Mistake Theory is cooperating, Conflict Theory is defecting.  If everyone around you is defecting, then it's naive and pointless to just cooperate and let them destroy you.  But if you can negotiate a mutually cooperating equilibrium, then it turns out better for everyone.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 10, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
It is accusing and treating a person as a defector. Which can be done independently of whether they are or not.  And, unless you are extremely kindhearted and forgiving, is typically followed by defecting against them.  If you're right and they are a defector, then you are justly retaliating against them.  If you're wrong and they were a cooperator you are unjustly defecting against them.  Either way, you're defecting.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Friday Fun Thread for October 14, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I've played Trails in the Sky 1 and 2, and about a third of 3 before getting bored of it, since it's very different from the first two.  I haven't played the other series in the franchise, since I was originally planning to play them all in order, but at this point I don't think I'm going to go back and finish trails in the sky 3.  If I do skip ahead, where do you recommend as a new entry point?


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 17, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
The first is a question of moral good, either utilitarian or deontological argument in nature.  Either the imposition on the mother causes greater moral harm in the utilitarian calculus than the moral harm of killing the fetus, or it deontologically takes a higher precendence than the fetus' life.


The second is a question of jurisdiction.  The claim is that it doesn't matter whether you believe the abortion itself is morally right or wrong, it's not you or the government's job to impose your view of morality on someone else.


So it's sort of a distinction on meta levels.  Claim 1 is that both abortion and not-abortion are morally acceptable in some semi-objective sense, so the woman to have the object level choice of abortion as people do in any choice when both options are acceptable.  Claim 2 is that morality is subjective, so each woman can decide for herself whether abortion is wrong or not, and then decide to do it or not based on her own internal morality.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 17, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I wonder if a lot of bullshit jobs are acting as a sort of... I'm not even sure what word to use here... emotional/comfort embezzlement?  A principal agent problem in which Middle managers, or even upper management and CEOs with plenty of slack that aren't worried about shareholders, hire or keep around people who make their work lives slightly more pleasant because that maximizes their own utility, not the company's.  If a middle manager has a choice between a friendly face who hangs out around the office, chats with them, brings donuts to work, and is generally friendly, or a stoic competent worker drone who gets stuff done, the optimal choice for the business is the worker drone, but the rationally selfish choice for the manager is going to depend on how much slack they have and how important it is for them to maximize productivity.  If the choice is between the friendly face in a bullshit job or removing the job and saving the company $50k, the rationally selfish choice is probably to the keep the friendly face around.  Having an extra friend around at the office isn't worth $50k, and the manager wouldn't spend that out of their own pocket.  But if it's just a number of a spreadsheet?  That's worth paying.  If the manager could take $50k of the company's money at put it into their own pocket, and get away with it, they would.  But they can't.  But spending $50k of the company's money to improve their lives and make their job slightly easier at an equivalent of $5k?  If they can get away with it then it's worth it, because the cost is measured in company-utility and the benefit is in personal-utility.


Any time you give someone control over someone else's money, inefficiencies are bound to happen in one way or another.  It's just a question of to what degree.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 24, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I wonder if some sort of mutually beneficial arrangement could be made between male and female homosexual pairs via polyamory.  That is, you have two men who are together romantically/sexually, two women who are together romantically/sexually, and the four cohabitate and share household responsibilities and raise a family together.  With six pairwise relationships there would be a lot more room for complicated drama, but they would have more slack when dividing labor and responsibilities, and more financial flexibility due to up to four incomes in a single household (or have one stay at home parent but still have 3/4 of the adults working jobs and earning money).


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 24, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I think the distinction would be that hereditarianism mostly rules out the possibility of an intervention that could work even if it hasn't been discovered yet, while environmentalism would suggest that better or more extreme interventions would work (in the most extreme case, baby swapping, though I don't think any serious environmentalists actually advocate this).  It's like the distinction between a mathematical theorem that has had no counter example found but is still open, versus a mathematical theorem that has been proven/disproven.


In the quantum mechanics case, there is no practical testable distinction and never will be, but in the HBD case there's the potential for a distinction in the future even if there's no meaningful difference for most people in the short-term.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				In considering runaway AGI scenarios, is Terminator all that inaccurate?
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Self preservation and resource gathering are subgoals that are highly conducive to accomplish other goals.  If the AI is destroyed, the AI cannot act, and its goals are unlikely to be fulfilled by chance.  Even if humans still want to accomplish the same goal as the AI, if the AI thinks it's better at accomplishing that goal then it will want to survive, unless the only means of survival directly thwart its goals in the process.  Therefore for a broad class of goals, not literally every goal but an awful lot of them, agents will logically conclude that self-preservation is useful and/or necessary to accomplish their goal.


This all falls under the category of "poorly considered directives", but it's a subset of that problem, not a distinct problem.  So it's incorrect say that they don't care about self-preservation.  They are unlikely to care about it as a base-level preference, but they'll still care about it as much as they care about any other necessary prerequisite of their main goal.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				In considering runaway AGI scenarios, is Terminator all that inaccurate?
			
		

	

	

	

		
			

			

				

			

		

			


				

					
It would be hilarious if the world ends because a movie generating AI decided decided the best way to generate realistic looking movies was to create exciting and violent scenarios in real life and film them.  Or just did that because it's blindly imitating things that it thinks movie directors do and doesn't notice the distinction between actors pretending to die and actually dying.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 31, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Has anyone at all been unbanned yet?  The most principled thing to do would be to create a formal, objective process for deciding who gets unbanned, and then the Trump decision logically follows from that process.  Therefore, Musk wouldn't actually know the answer until the process has been determined.  And maybe the Trump question is a big enough deal that it would bias the decision of how the process would work (maybe carving out exceptions to make it go one way or another), or maybe it doesn't.  I expect that most reasonable processes would end up unbanning Trump, but I guess it depends on the criteria (maybe potential unbanees need to jump through some hoops to demonstrate the unjustness of their ban, and Trump doesn't feel like going through that).


I have no idea of Musk is actually going to do things that way.  But if he is then there's like a 90% chance that Trump gets unbanned, but it isn't guranteed one way or another, so of course Musk can't commit to an answer yet.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 31, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I'm not sure that it does.  A good process would be one which treats people fairly and simultaneously handles large people well.  If Catturd advocates for violence but the only obstacle to actually causing it is his lack of audience, then he should be banned for it the same as a larger person would be.  If Catturd is allowed to say something mean but nonviolent, then someone large like Trump should be able to say the same mean but nonviolent things.  A good set of rules is not consequentialist, because that's impossible to practically enforce since no one can predict the future, which just leads to subjective favoritism, as we've seen on Twitter so far.  A good set of rules should be egalitarian, deontological, and as objective as reasonably possible so that people can predict ahead of time what is and is not allowed and then choose to act accordingly.


The class problems in our society are not caused by influential people being held to the same standard as everyone else when they need higher standards.  The problem is that the standards aren't being enforced equally or objectively and so in effect they tend to be held to lower standards and find loopholes.  Trump should be held to exactly the same standards as everyone else, because fairness is an important principle that we need more of in our society, especially from internet platforms.  And if Trump being held to those standards would lead to a bad outcome then create a better set of standards that leads to a good outcome and hold everyone to those instead.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 24, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I guess that's a good point.  So really it means hereditarianism rules out the possibility of behavioral interventions on the individual level but allows for genetic/eugenic interventions on the population level.  Which are less useful given you can't apply it to already existing people, and generally less tasteful to most people, and harder to enact ethically.  But theoretically tenable if you can pinpoint the actual IQ genes.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 31, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I like this theory, although I would describe his evidence as "slightly flawed" rather than "total bunk".  The Baratheon family all having black hair implies that both they and the people they tend to marry mostly have two black hair genes rather than one, otherwise you'd expect 1 in 4 of them to be lighter.  This doesn't prohibit the case that one of Robert's parents might have had a recessive light hair gene and passed it on to him, but it does make it less likely.


Further, Ned doesn't learn of the family history and immediately falsely conclude that it logically implies Joffrey is a bastard, but it does raise the possibility.  He then proceeds to check Robert's bastards, which again all have dark hair.  If Robert did have a recessive light hair gene, you might expect some of his children to have light hair, but they do not.  Again, this does not logically preclude the possibility, but it makes it statistically less likely.  Combine this with the fact that all three of his supposed children are light haired, and even in the case he has a recessive gene there's only a 1/8 chance all 3 children would get it.  Combining all of these is a lot of statistical evidence, even if it's not a logical certainty.


And of course, Ned doesn't actually understand genetics or statistics, but he does understand vague ideas about heredity and combining multiple sources of evidence, so I think his general process is heuristically valid.  And again, he doesn't take it as logical certainty, he takes the evidence and confronts Cersei, who then confesses.  Which is probably a mistake politically as it ends up getting him killed, but not a mistake in terms of deduction given the information available to him.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of October 31, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
This reminds me of a quality contribution post, regarding the "it's okay to be white posters", which says  "The far-right statement isn't: "It's okay to be white" , it is you tearing it down."


link: https://www.vault.themotte.org/post/the_shortest_quality_contribution


Part of why trolling sometimes works as an ideological tool is because people who react disproportionately to innocuous things look ridiculous and usually end up hurting their own side more than they help.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Don't even go there: NIH blocks access to genetics database
			
		

	

	

	

		
			

			

				

			

		

			


				

					
How does access to the database work?  Do scientists request specific types of data which are sent to them if their proposal is approved?  Do they get temporary access keys to look stuff up?


I'm wondering how plausible it would be for someone to pirate/copy the database and replicate it elsewhere anonymously kind of like sci-hub.  Which would be more or less plausible depending on who has how much access to the database.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Don't even go there: NIH blocks access to genetics database
			
		

	

	

	

		
			

			

				

			

		

			


				

					


Access is by study or study subset and so would require a massive distributed project of requests that were falsified or a big internal wikileaks style breach. That would really suck for all the patients who were generous enough to share their data with the assumption it would be protected. I think it is a Bad Thing to be avoided.




Probably.  In addition to the damage to those specific individuals, it would make it significantly harder to convince future patients to commit to similar projects.  But on the other hand, it would reduce barriers to scientific progress and the authoritarian control of elite institutions from being able to arbitrate which topics are and are not within the Overton Window of Science.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of November 7, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I feel like this might run into issues with friction, where people don't consciously want to go through the effort of paying for content they don't yet know if they'll like.  An alternative would be some sort of bundled subscription service like Netflix/Hulu/Etc.  If you could just pay a fixed rate for access to most premium content on the internet, then people could just sign up for one thing and not worry about it.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of November 7, 2022
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Look with your eyes.  If you see someone, you can usually tell what race they are.  If race had no biological basis, you wouldn't be able to do this because there would not be reliable clusters of traits that you could identify.  They aren't perfectly defined clusters with sharp boundaries: you cannot with 100% accuracy identify the race of every person just by looking at them, and there are patterns which go beyond what the eye can visibly determine.  And it's largely subjective: human-defined categories that decide which patterns do or do not count as part of which race.  But the foundation is clearly biological otherwise there wouldn't be any patterns for you to detect in the first place.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Don't even go there: NIH blocks access to genetics database
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I assume a theoretical leaker would leak anonymously, but I guess if the data set is unique to that study then they could deduce it, unless a bunch of them were combined and mixed together, maybe with some stochastic omissions to further obfuscate what the original data looked like.  A deadman's switch might work, where the data gets uploaded to the internet and made public like 10 years later.


But you're right that there would also be the issue that nobody could publish results using the leaked data.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Does my Philosophy of Sexuality Professor Have a Point? (It's a mandatory gen-ed)
			
		

	

	

	

		
			

			

				

			

		

			


				

					
We have no obligations on our selection of friends or romantic partners by any criteria.  It is your freedom as a human being to make decisions regarding yourself and your relationships any way you want.  You can choose to only befriend people wearing orange hats, or only people with only one arm, or only people whose name starts with the letter B, and that's your right.  It would be silly, but you can do that and nobody has the right to force you to befriend people according to some other set of criteria.


Now, importantly, if you actively dislike people for weird criteria that might make you an asshole.  That is, if you think people whose name starts with A or C-Z are inherently evil and that's the reason you won't befriend them then you're doing something wrong.  If you befriend everyone but play favorites with your B friends and give them gifts while badmouthing your other friends then you're doing something wrong.  But if you simply decide that you want to have a gathering of only B friends because it sounds like fun, that's entirely within your rights to do.  It's stupid, but you're allowed to do stupid things.  You should not drop people below neutral due to race or sex or other immutable characteristics.  But you can choose not to elevate them above it.


This is especially true regarding romantic partners.  Nobody has the right to force or compel or pressure you into dating someone you do not actively want to date.


Therefore, all premises and conclusions are false.