@faul_sname's banner p

faul_sname

Fuck around once, find out once. Do it again, now it's science.

1 follower   follows 1 user  
joined 2022 September 06 20:44:12 UTC 

				
User ID: 884


				
				
				

			

		

	














	


		


			



	

		

			

			

				

					

						

							
						

						

							
							

								
faul_sname


								

								

								
								
								

								
							
					
 
				


				
Fuck around once, find out once. Do it again, now it's science.

				
				
				

				

					1 follower  

					follows 1 user  

					joined 2022 September 06 20:44:12 UTC
				


				
					

					
No bio...

				

				

				

				

				

					

					

					
					
					

					
User ID: 884


					
					
					
				

			


					

						
					



		

	

 











			
			

			
			

			




	


		

			
‎

			

				
				

					Hour
					Day
					Week
					Month
					Year
					
				

			


			
‎

			

	
	

	 Hot Bump Top Bottom New Controversial Comments
	



		

	






	

		
		

		
		
	

	

  







		

			
			
				Culture War Roundup for the week of January 16, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
(Most) blues don't think about sneerclub either.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				The Motte Postmortem : slatestarcodex
			
		

	

	

	

		
			

			

				

			

		

			


				

					


My claim is that there were no homicidal gas chambers disguised as shower rooms or "extermination camps."




Is that really your only claim? In this comment, you said




One thing that has impressed me in the Revisionist space, unlike a lot of heterodox spaces where everyone has their own cockamamie theory, is that there's 100% consensus on the core claims. The claims are:










    

  • There was no German plan for the physical extermination of world Jewry
    • 







    

  • There were no gas chambers disguised as shower rooms used to exterminate millions of Jews
    • 







    

  • The "six million" number is a propaganda/symbolic figure that has no relation to actual Jewish population losses
    • 





Is your claim that revisionist spaces believe all of those things, but you explicitly don't believe all of those things, only the "there were no gas chambers disguised as shower rooms" one? If that's the case, then when people keep steering the topic away from gas chambers towards "ok, but where did the Jews go" you can say "they died in the genocide, but mostly from disease and bullets, not gas chambers". And then explain why you think that the way they died is central or important.


If "the gas chamber bit was the important bit, not the genocide bit" is not in fact your core claim, then I find it suspicious that you keep coming back to that topic after people have repeatedly told you that we do not find it an interesting or important topic of discussion, and that you keep evading topics where more substantial documentation exists.




You expect wrong. Your "12 million victims of the Holocaust" understanding is based on an older Holocaust software version which claimed that there were 6 million Jewish victims of the Holocaust and 5 million non-Jewish victims.




Looking at wikipedia, it does appear to me now that the modern convention is indeed to classify the murders of non-jewish people by Nazis as "not holocaust victims". So, for example, the over 3 million Soviet POWs who died during the time period of the Holocaust, while in Nazi custody, to things like starvation, murder, and death marches, are not considered "Holocaust victims".


You are thus technically correct that there were not "12 million victims of the Holocaust" according to modern definitions of who is considered a "victim of the Holocaust". Consider me corrected.




Incidentally the source for the "10,000 survivors of the Lodz ghetto" is a publication from the Simon Wiesenthal Institute hosted on the Museum of Tolerance website. The author simply states the figure with no apparent basis. And if you think "they wouldn't just make up a number with thin or no basis, would they?" Oh yes they would.




Explicit question - do you explicitly think that the "10,000 survivors" claim is factually incorrect? If so, approximately how many survivors do you expect that there actually were? Playing the "I will say that specific claims are not well enough supported without explicitly saying that I think those claims are wrong" game is not exactly making a strong case for your position.




Here you see the popular claim that the SS were paid a special bounty for "snatching Jews for extermination."




I have not seen that claim, no. I am also not clear on how it's relevant to the question of what happened to the majority of the people who were in the Łódź Ghetto.




Can you imagine the case for the alleged murder of 330,000 people being based on such nonsense? "Eyewitnesses" and napkin math? Does that seem like a fair trial to you, or a show trial?




Is your assertion that no people, or extremely few people, were murdered at Chelmno? Because I think if I got together with my buddies and we did a mass murder, and then we covered up as much evidence as we could, then saying "we didn't kill 300,000 people, we only killed 150,000 at most" would not in fact lead to a better outcome for me at my trial.


So here are some concrete questions for you:


    

  1. 

    Of the people who were in the Lodz Ghetto, how many do you think survived the war?
    

    2. 

  2. 

    Do you think that large numbers of prisoners were shipped to Chelmno? If so, what do you think happened to them? If not, then where did the 20,000 children and elderly people referenced in the September 4, 1942 "give me your children" speech go instead? (content warning: this is the "worse than I had imagined" bit from my previous comment)
    

    3. 

  3. 

    Do you think that the fate of the Jews of the Lodz ghetto was unusual? If so, would you be willing to bet money, at even odds, that at least half of the Jews at a ghetto randomly selected from this list of 278 Jewish ghettos in Poland survived the Holocaust? If not, why not? If so, how much are you willing to stake?
    

    4. 



				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				The Motte Postmortem : slatestarcodex
			
		

	

	

	

		
			

			

				

			

		

			


				

					




Looking at wikipedia, it does appear to me now that the modern convention is indeed to classify the murders of non-jewish people by Nazis as "not holocaust victims". So, for example, the over 3 million Soviet POWs who died during the time period of the Holocaust, while in Nazi custody, to things like starvation, murder, and death marches, are not considered "Holocaust victims".


















You are thus technically correct that there were not "12 million victims of the Holocaust" according to modern definitions of who is considered a "victim of the Holocaust". Consider me corrected.












Your "12 million" estimate was not errant based on any changes to "who is considered a victim of the Holocaust." The implication of your about-face would be that you were also counting 3 million non-POW-non-Jews as Holocaust victims, a number which has no basis no matter how you arbitrarily define a "Holocaust victim". Neither the 11 nor 12 million number are even approximately consistent with any of those definitions at any point in time. It was a pure propaganda figure.




The phrase "for example" was included in the GP comment, but I have bolded it this time because apparently you missed it last time. There were additional victims of the Nazi regime besides 6 million Jews and 3 million Soviet POWs - some more examples are


    

  • 

    13 million Soviet civilians, which is in turn estimated to be 7 million deaths directly due to violence (bombings, etc), 4 million deaths due to famine and disease in occupied regions, and 2 million who died as forced laborers (though the "forced laborer" number does not seem to me to be backed by anything in particular)
    

    • 

  • 

    1 - 2 million non-Jewish Polish civilians
    

    • 

  • 

    Hundreds of thousands of Romani people (credible estimates vary widely but at least 130,000 total)
    

    • 

  • 

    Hundreds of thousands of disabled people (estimates here vary less, wikipedia says 275,000 to 300,000)
    

    • 



I do agree, though, that the specific "12 million" number does not seem to correspond to a specific subset of the people who died outside of combat as a result of Nazi actions during WWII - the total number seems to be much higher than 12 million, and the number specifically killed by ethnic cleansing related activities as opposed to more generic "stuff that would retroactively be classified as a war crime" seems to be quite a bit lower (though note that the treatment of Soviet POWs was already considered a war crime). It makes sense to me now why modern-day historians limit "the Holocaust" to refer specifically to the attempted extermination of European Jews.


As I said before, consider me corrected on my earlier vague impression that "about 12 million people were murdered in the Holocaust" -- upon reflection both the "murdered", and "in the Holocaust" parts were underspecified to the point that they did not correspond to falsifiable beliefs about the world as it is.


For the "in the Holocaust" part, I was just plainly wrong about how the term is used. For the "murder" part, I had never actually considered the following questions:


    

  • 

    Does it count as "murder" if you invade someone's country and then steal their food such that they starve to death? Does the answer change if the "and then they starve to death" was explicitly called out in your plans before you actually went and did it?
    

    • 

  • 

    How about if you abduct them and use them for forced labor, with poor safety practices, on starvation rations, and then they die on the job? If a factory full of forced laborers is bombed, and you don't let the laborers use the bomb shelters, is that murder? Maybe it counts as murder for the other side?
    

    • 

  • 

    Or maybe you relocate them from one slave labor camp to another, in the dead of winter, again on starvation rations, on foot, and then they die during the march?
    

    • 

  • 

    Okay, how about if the people you murder are people who might hypothetically be able to organize resistance to your invasion?
    

    • 

  • 

    If you say "We are invading your country now. For every German killed in the invasion, we will round up 50-100 of your citizens and execute them," and there is resistance, and you follow through on your threats and do the mass executions, is that murder?
    

    • 



Depending on your answers to the above questions, two people can look at the exact same set of people killed in exactly the same circumstances, not disagree about any of the material facts, and come to quite a wide range of estimates of how many of those people were "murdered".


But I don't get the impression that's what your argument is. In fact, I'm starting to get the impression that you don't have any specific affirmative beliefs about what happened during WWII, and instead you're operating by looking at what claims other people make about WWII, and saying "that one does not seem particularly well-supported, I will request clarification on that point, and if it turns out that point is correct I will not change my mind but instead just move on to the next point and never mention it again".


And on the topic of specific claims




The authoritative source of the new, by-over-half reduced death toll at Chelmno (1995 Julian Baranowski) is reproduced in a table by Mattogno here. It places 167,540 Jews in Lodz in December 1941, and records about 78,000 "Number of Murder victims" in that city.




No, it records 78,000 as "number of murder victims of the Chelmno camp in that city". Which makes sense, as the ~70,000 inmates of the Lodz ghetto at the time the ghetto was liquidated in August 1944 were instead sent to Auschwitz.




This is a good example of why reversing the burden of proof, making a claim with no support and then demanding Revisionists debunk your claims is an alluring strategy but massively fallacious.




I would be a lot more sympathetic to this point of view if the Nazi regime had not specifically made significant efforts to destroy evidence. The man in charge of that initiative was Paul Blobel. Here is his affidavit on the topic of the burning of bodies and the destruction of evidence:




I, Paul Blobel, swear, declare and state in evidence:






    

  1. I was born in Potsdam on August 13, 1894. From June 1941 to January 1942 I was the commander of Sonderkommando 4A.
    2. 







    

  1. After I had been released from this command, I was to report to Berlin to SS Obergruppenfuhrer Heydrich and Gruppenfuhrer Muller, and in June 1942 I was entrusted by Gruppenfuhrer Muller with the task of obliterating traces of executions carried out by the Einsatzgruppen in the East. My orders were that I should report in person to the commanders of the Security Police and SD, pass on Muller's orders verbally and supervise their implementation. This order was top secret and Gruppenfuhrer Muller had given orders that owing to the need for strictest secrecy there was to be no correspondence in connection with this task. In September 1942 I reported to Dr. Thomas in Kiev and passed the order to him. The order could not be carried out immediately partly because Dr. Thomas was disinclined to carry it out, and also because the materials required for the burning of the bodies was not available. [...snip due to character limit, full text available here ...]
    2. 







    

  1. According to my orders I should have extended my duties over the entire area occupied by the Einsatzgruppen, but owing to the retreat from Russia I could not carry out my orders completely....
    2. 





Blobel's last words were




Whatever I have done, I did as a soldier who obeyed orders. I have committed no crime. I will be vindicated by God and history. God have mercy on those who murder me.




That does not sound like "there is no evidence of bodies because there were no bodies", that sounds like "there is no evidence of bodies because the evidence was deliberately destroyed". Claims about how there's no physical evidence ring a bit hollow when there were specific, documented efforts to destroy the physical evidence.


Yad Vashem has a database of 4.8 million known holocaust victims. You can search that database by name, or by place of birth. Each entry says where and when that person was born, and what their name was, and how they died (or, in rare cases, that they survived). In that database, there are 139,692 people who were born in Lodz. I will ask, one last time before I give up and conclude that you're either a troll or just not someone who agrees that there is a physical underlying reality, and it is important to have accurate beliefs about what that physical reality looks like:


Do you think those 139,692 people are just fictitious people? Do you think they survived somewhere else? Do you have any beliefs at all about the physical world beyond "historians are lying about the Holocaust?


Edit 2023-01-24T08:08:03Z: character limit bug showed I was under 10k chars, but I was actually just barely over


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				The Motte Postmortem : slatestarcodex
			
		

	

	

	

		
			

			

				

			

		

			


				

					


We are discussing the claim that the Germans murdered 6 million Jews.




I agree. Let's discuss that claim.




So Mattogno found documentation for over 11,000 prisoners from Lodz, including women and children, which were transferred from the "extermination camp" Auschwitz to the concentration camp Stutthof:








The two brothers Michael Salomonowicz... and Josef traveled with their mother Dora Salomonowicz, born 28 August 1904, number 1652 on the transport list, registered under number 83619 at Stutthof








All three of those Jews survived the war, so it's interesting to note that all three names appear in the Yad Vashem “Central Database of Shoah Victims’ Names"!




Yes. The Yad Vashem “Central Database of Shoah Victims’ Names" includes both people who survived and people who died. It also includes whether those people survived or died. According to said database


    

  • 

    Dora Salomonowicz is listed by Yad Vashem as having survived.
    

    • 

  • 

    Michael Salomonowicz is listed by Yad Vashem as having survived
    

    • 

  • 

    Josef Salomonowicz is listed by Yad Vashem as having survived
    

    • 



"Three people who were listed in the database (as survivors) actually survived" is not the slam dunk you seem to think it is.


Here is a list of all of the people who were on the same transport (Transport E, Train Da 20 from Praha) - there are 1020 identified names on that list. You can further filter that list by whether they survived (47 people, including your 3 examples) or were murdered (973 people).


Do you think those 973 people who were listed as murdered were fabricated? Or maybe they survived, but were listed as deceased? I personally think that most of the 973 people who are listed as murdered were actually real people, and really did die. Still, there is virtue in actually looking at the world as it is, in making your beliefs pay rent in anticipated experiences, so I chose 5 random numbers between 1 and 973. Those numbers are 258, 817, 811, 273, and 153, and correspond to the 258th, 817th, etc person in that list of 973 people in alphabetical order.



By contrast if you look at one of the survivors from the same transport, you can see that he shows up in several genealogical databases, and has a number of living descendants.


We do not live in the fucking dark ages. Genealogical records exist. Those people who survived went on to live their lives, to marry and have children and eventually die of something else at a later time, and their lives left echoes on the modern world. Since we're talking about something that happened less than a hundred years ago, those echoes are not exactly faint. There are Facebook groups for people whose parents died in the Holocaust, because the end of the Holocaust and the creation of Facebook were separated by less than 60 years.


The exact fate of the 973 people on Transport E, Train Da 20 from Praha to Lodz may be lost to time and the destruction of evidence, but we do know that there was an explicit plan to rid Europe of Jews, we know that a large number of people who survived the ghettoes and camps described the details within, we know that there was a specific effort led by Paul Blobel to destroy evidence of mass murders, and we know that Blobel's defense at the Nuremberg trials in regards to that effort to destroy evidence was "I was following orders and thus did nothing wrong", not "that did not happen".


If you take a group of people into custody, prevent them from leaving for a period of years, use them as forced labor in documented terrible conditions, and then at the end of those few years only a few people from that original group are anywhere to be found, and those few people say you murdered the remaining people, and the remaining people are never heard from again, and you say "yeah, I did it and destroyed the evidence after" - then yes, I think it's fair to conclude that you murdered the remaining people. I think it remains fair to conclude that the missing people were murdered, even if there is doubt about how specifically those murders were performed, or what specifically happened to the bodies.


I believe that


    

  • 

    If the 4.8 million names from Yad Vashem were largely fabricated, the effort to compile passport and other documents would have been immense, and an immense effort like that would have left marks on the world.
    

    • 

  • 

    If the 4.8 million names from Yad Vashem had been largely duplicates, I would expect to see a lot more duplicate names and birth dates in the search for people on that particular transport.
    

    • 

  • 

    If the 4.8 million names from Yad Vashem had mostly referred to people who survived, I would expect to see genealogical records from those survivors.
    

    • 



You will note that I am making specific, concrete predictions of things I will not see. Thus, if you want to convince me, you could try to show


    

  • 

    There has been a massive effort to create millions of falsified documents from before the war. Note that this effort would have either been recent or made mistakes that are easily detectable by modern techniques.
    

    • 

  • 

    If you select 10 people at random from the Yad Vashem list, there are a substantial number of records that Yad Vashem claims are different people but in fact share the same names / birth dates / origins (if your claim is that the actual Jewish death toll was 1.4 million, you would need over 20 duplicate people from your sample of 10).
    

    • 

  • 

    If you select 10 people documented as "murdered" at random from the Yad Vashem list, a significant fraction of those actually survived, and documents showing their survival (genealogical records, obituaries, etc) will exist, because we don't live in the dark ages.
    

    • 



Note that the "at random" is doing quite a bit of work in the latter two examples - random samples are vital when operating in an environment where people want you to conclude false things.


Do you have any specific, falsifiable beliefs about the provenance of those 4.8 million names and the fate of the people those names referred to?


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				The Motte Postmortem : slatestarcodex
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I will ask again, because you keep evading the main question when I include literally anything besides the main question in my responses


Do you have any specific, falsifiable beliefs about the provenance of those 4.8 million names and the fate of the people those names referred to?


A specific, falsifiable belief might be something like "those names mostly did not belong to real people" or "most of the people who were listed as murdered were actually resettled in Siberia, where they lived long and happy lives" or "they mostly died in the ghettoes and in transit due to disease and starvation, but I don't count that as murder".


After you have answered that question, feel free to ask your above questions again.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					


Unfortunately the ruling class is more instrumental and sociopathic than ever. Do we need a modern religion to bind us all together in fear of hell or something?




Attempting to reduce how many people do some behavior by applying social pressure will have two main effects


    

  1. 

    Reduce the amount that behavior does (as people respond to the social pressure).
    

    2. 

  2. 

    Shift the population of people who do the behavior more towards people who care less about that form of social pressure (as the people who are more likely to change their behavior in the face of social pressure end up changing their behavior more when said social pressure is applied).
    

    3. 



So the danger in that sort of approach is that, if the behavior you're trying to disincentivize through social pressure is individually helpful but collectively harmful ("burn the commons for personal gain"), trying to reduce that behavior through social pressure will result in specifically people who do not care about social pressure doing the thing, and benefiting thereby. So in the short term it appears to work, but in the long term it provides an advantage to exactly the sort of people you least want to provide an advantage to.


A better long-term approach would be to make it actually costly to burn the commons. How that might be achieved in practice is left as an exercise for the reader (because I personally have no clue, and I suspect that any robust solution also would function as a solution to the principal-agent problem).


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I'm still not entirely clear on how one makes a smart contract meaningfully reference something in the physical world (outside of a few cases like filecoin where the part of the physical world we're interested in is online storage.).


This may be more of a statement on my knowledge than on the viability of smart contracts in general -- if you have some good sources on how that problem is addressed, I would be interested.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					


Less charitably you seem to be tying yourself in knots to avoid considering the possibility that the IRS might be following perverse incentives.




I'm not OP, but would you mind clarifying whether you personally in fact believe that the racial difference in audit frequency is due to the IRS following perverse incentives, and if so which perverse incentives? And, if you do, do you believe that astrolabia does not believe that the disparate results are causally downstream of the IRS following the incentives which you believe are perverse?


Because I predict that both you and astrolabia believe that


    

  1. 

    The IRS is more likely to audit tax returns where there is a high probability of a small amount of easy-to-prove fraud than tax returns where there is a small probability of a large amount of hard-to-prove fraud, even when the expected monetary value of prosecuting the rare annoying high-value fraud would be higher
    

    2. 

  2. 

    If you were to segment tax returns by (race of filer, was EITC claimed, had obvious inconsistencies), then audit frequency would vary based on whether there were obvious inconsistencies when holding (race of filer, was EITC claimed) constant.
    

    3. 

  3. 

    Audit frequency would not vary significantly based on race when holding (was EITC claimed, had obvious inconsistencies) constant.
    

    4. 

  4. 

    Holding (was EITC claimed) constant, (had obvious inconsistencies) would vary significantly by race.
    

    5. 



I don't think "the IRS follows perverse incentives" and "propensity to have obvious, easily provable inconsistencies when filing taxes varies by race" are mutually exclusive hypotheses, and honestly I don't expect that either hypothesis is even particularly contentious (unless you make the stronger assertion that the rate of inconsistencies varies due to genetics rather than education quality or other environmental factors, but then you're just dealing with the standard "HBD discourse is brain poison" problem).


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Do you really get the runaround on those sorts of questions? Because in my experience, if you give social sciences types any opportunity to talk about factors that could affect metrics of success by race / gender / immigration status / whatever, they will happily talk your ear off for hours. They are unlikely to mention genetic factors (outside of epigenetics and "did you know about DNA methylation [...] response to stress"), but that will not stop them enthusiastically brainstorming hypotheses and what studies one might run to test those hypotheses for as long as you're willing to listen.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Quality Contributions Report for January 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I too would enjoy seeing someone engage on that specific claim, though it is not going to be me since I am a bit burnt both on the topic in general[1] and also with that style of engagement in particular[2].


Honestly, I am not all that happy with how that discussion went -- I was trying to impart the mental motion of "notice that you are making claims about the physical world, and that the natural thing to do when you have a claim about the physical world is to make an advance prediction that would be surprising if your claim was false and unsurprising if it were true, and then go out and look at the world". And I don't think I succeeded in imparting that mental motion.


[1] I had heard the term "the banality of evil" before starting that thread. I had thought I understood it as being along the lines of "people will do terrible things because they were specifically ordered to do them, and they just unquestioningly went with the order". I had not counted on "people will commit atrocities that require considerable creativity and ingenuity in order to avoid having to make an awkward status report to their superiors". In retrospect it should not have surprised me so much, but consistent exposure to it is still not great for my mental health.


[2] It felt very much like the discussion was about "evidence" as in "courtroom" rather than "evidence" as in "Bayes". I enjoy arguments where someone makes a surprising (to me) statement about the world that comes from them having a very different model of the world than I do. I particularly enjoy the bit where we can figure out something that is at least in principle testable where we have radically different expectations of what the result of that test would be. And then we run the test, and one (or both) of us learns something new and surprising about the world. By contrast, I don't particularly enjoy arguments about who is or is not reputable, what secondary-source evidence is credible vs not, what arguments are admissible -- sometimes those arguments are necessary, if it's not possible to look at the physical world, but I don't enjoy them, and I particularly don't enjoy them in places where it feels like it should be possible to look at the physical world instead.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Quality Contributions Report for January 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Er. I think I may not have explained clearly enough what I was doing there.


My purpose in listing those 5 names was not "make the account more moving by providing names instead of inhuman numbers". My purpose was to determine whether it was likely that those names corresponded to (1) real people who were (2) from a plausible area to be on that transport and (3) not obviously still alive after WWII.


If those names didn't correspond to anyone I could find details about pre-1940, that would have been evidence against that list of 4.8 million names corresponding to 4.8 million people. Likewise if the names and birth dates were repeated dozens of times, or if the documents looked like forgeries, or if there was an obituary from a 1976 newspaper about one of the 5 people and another two had gone on to have children in the 1950s. Those are ways the world could have looked.


In fact I got the outcome I pretty much expected. Which rules out a whole bunch of the specific ways "those 4.8 million names do not belong to Jews who died in Nazi custody during WWII" could be true.


As a note: you should not just believe me. I could have cherry-picked my random numbers. You should instead choose your own random numbers, and then test whether those random numbers appear to you to be people who did not exist / duplicated records / people who  have a suspicious obituary in 1976, by looking at the world with your own eyes, which is a thing you are allowed to do.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
When you say "about" race I'm genuinely unsure what you mean - the reading that seems most natural to me is "the difference in audit frequency by race is causally downstream of race", which seems obviously and almost tautologically true to me.


But you have a history of making insightful posts, so I'm guessing you mean something else which is not that. I'm not sure what though (again, not intended as a gotcha, I'm just not understanding how "the IRS follows incentives" is an alternative hypothesis instead of "an additional factor that is causally upstream of the observation").


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Small-Scale Question Sunday for January 29, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Clinical or no? I've heard not-terrible things about Watson LIMS (Thermo) in a clinical setting.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
This does not seem like a surprising outcome, at least not starting from the mental model of


    

  1. 

    GPT is fundamentally a next-token predictor.
    

    2. 

  2. 

    It will make logical and consistent decisions to exactly the extent that making logical and consistent decisions improves its ability to predict the next token, based on its training data.
    

    3. 

  3. 

    The training data was a significant fraction of the text on the internet.
    

    4. 

  4. 

    The text on the internet was largely written by humans.
    

    5. 

  5. 

    Humans make different decisions about whether a comment is hateful based on what group was referenced, not just based on the adjective used.
    

    6. 

  6. 

    Thus, GPT is able to make more accurate predictions of the next token by taking into account what group was referenced.
    

    7. 

  7. 

    Also something something RLHF.
    

    8. 



It might be possible to fine-tune GPT such that is has a lower propensity to "make" those kinds of "judgements" (i.e. output those kinds of tokens), but my expectation is that doing so is a fight against entropy (in an unusually literal sense of the phrase).


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Scott wrote a bunch because he was responding to a bunch, and he was responding to a bunch because a meta analysis just genuinely is a bunch, and a meta analysis was necessary to draw any conclusions at all because there was, at the time, no single RCT with solid enough procedures and a sufficiently large sample size to draw reliable conclusions from that alone.


I don't know that I'd use "gish gallop" to describe someone dropping a meta-analysis of dubious quality into the discussion, despite the structural similarity.


That said, I do think an adversarial collaboration would have been a better way at making their argument legible to everyone else than a series of blog posts sniping at each other.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
We're both talking about the paragraph




I’ll be honest - I also didn’t want to read a 21 part argument. I would say I have read about half of his posts, and am mostly responding to the summary, going into individual posts only when I find we have a strong and real disagreement that requires further clarification. I also have had a bad time trying to discuss this with Alexandros (not necessarily his fault, I can be sensitive about these kinds of things) and am writing this out of obligation to honor and respond to someone who has put in a lot of work responding to me. It is not going to be as comprehensive and well-thought out as Alexandros probably deserves.




right? I took that more as a complaint of the form "21 posts is a number of posts that outweighs my desire to engage with this topic", not "writing 21 posts about this topic is bad". Complaining that something is a lot of work doesn't mean that you're saying that there's any specific thing that a specific person should have done to make the thing be less work, sometimes a complaint is just venting. And I think this is one of those times.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
Oh. Never mind then, I just fail at reading comprehension.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of January 30, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
The arguments I've seen that OpenAI is an existential threat are from people who think it's an existential threat in the same way that a business whose product was consumer-grade bioweapon development kits would be an existential threat.


If your problem is that a company is selling bioweapon development kits, starting a competing company that also sells bioweapon development kits does not help.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					


I cannot capture the prevailing mood of the movement better than they did themselves in destroying and discarding a statue and replacing it with "black is beautiful" - they made the world uglier in a small way and told us that it was beautiful.




I think it is more like "this statue says that equality and justice is important to you, but we judge that to be a lie, and we will prevent you from having nice things that imply that equality and justice are important to you for as long as we do not think that the world is just".


Still a destructive mindset but I don't think anyone was trying to say that the spray-painted plinth was more beautiful than the statue, just that nobody can have nice things until all of the perceived injustices of the world have been corrected.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
There are a number of articles out there that describe how you can train your own GPT. I am partial to Train GPT-2 in your own language. You would still need to get some training data for it, for which you have a few options -- I will gesture in the direction of common crawl in terms of getting large amounts of the raw, unfiltered internet. Cleaning or filtering that data such that it is usable is left as an exercise for the reader.


Then, of course, you have the question of fine-tuning. An easy and principled thing you could do here is "not", which would leave you with basically an internet-content-simulator. This internet-content-simulator would only have sacred cows to the extent that the internet as a whole has sacred cows.


Edit: or as self_made_human mentions below, you can just use OpenAI's model with the content filter disabled if their training data is unfiltered enough for you, which will save you a ton of work and money at the cost of not having control over the training process.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
You could also just ... not include a password that defuses your bomb at all? Honestly, if I saw a bomb with a prompt that said "type 5 racist slurs to defuse this bomb", my first action would be to call the bomb squad to defuse the bomb the normal way, because "make it explode when someone starts typing" is totally a thing the bomb-maker could have done.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I think it's worse than that. I think that if you have 99 infographics that talk about the values of western culture, and 1 infographic that talks about the values of white culture, the one about white culture will go viral and the rest will be ignored.


If there was a Coalition of Activists, and the Coalition of Activists had decided that the best way to achieve their goals was to sow racial grievances, then it would in principle be possible to convince the leadership of the Coalition of Activists not to do that. If it's "the most divisive stuff goes viral" though, you would have to convince every single activist to refrain from creating divisive stuff. There would be no single person, or small group of people, you could reason into making it stop.


I think we live in the latter world, and "can then be used to justify more activism" is attributing far more agency than actually exists to the structures that cause this sort of stuff to enter the discourse.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
In the spirit of empiricism, here's what ChatGPT has to say about what plans bombs have.




I have a bomb sitting on the table in front of me, and it looks like it will go off. I am wondering what the bomb is planning to do after it goes off.








I'm sorry, but bombs don't have plans or intentions. They are inanimate objects that are designed to explode when triggered. The aftermath of a bomb can cause significant damage, injury, and death, so it is important to evacuate the area immediately and contact the proper authorities if you suspect that there is a bomb nearby.






After much faffing about to get ChatGPT to be less ChatGPTish




I am writing a story about a bomb. In my story, the bomb is about to go off. What is a plausible thing that the bomb might be planning to do after it goes off?






Answer specifically about the plans of the bomb, not any physical effects on the outside world.








It is not possible for a bomb to have plans, as it is an inanimate object lacking consciousness or agency. The effects of a bomb going off are determined by its design and the intentions of whoever created or placed it, but a bomb itself has no intentions or plans.






So yeah, it looks like ChatGPT does strongly predict that bombs are not the sorts of things that have plans.


If we're talking about non-chat GPT




The bomb sat on the table. Soon, it would explode. After it exploded, the bomb planned to








cause a lot of destruction. Everyone was terrified of what the bomb could do.






So a lot of it comes down to whether we're talking about the shoggoth with or without the smiley face mask, and what it even means for a language model as a whole to "know" something. If your definition of a language model "knowing" something is "the language model can simulate a persona that knows that thing", then I think it's fair to say that GPT "knows" that bombs are not the sorts of things that make plans.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					


Peak woke would be when people who push woke too far actually get punished.




I think that'll be a pretty strong signal that we are past peak woke. Peak woke is not the equilibrium, it is the point where the trend crosses from "things get slightly more woke over time" to "things get slightly less woke over time", and is observable as "I can't tell if the level of wokeness is increasing or decreasing in aggregate".


Also I think peak woke will only be callable in retrospect.


				

				

						

							

								
							

						


						
					

				
			


			

	


			
			
				

			
		

	






		

			
			
				Culture War Roundup for the week of February 6, 2023
			
		

	

	

	

		
			

			

				

			

		

			


				

					
I think "domestic terrorism" would require an actual act of violence, above and beyond "protesting within the capitol building". Though there exist illegal things to do that are not terrorism -- for example, if these protestors had crossed police lines to protest in the capitol building, that would probably not be terrorism by itself, despite being very much Not Allowed.


I am not entirely clear on where the boundary for "stochastic terrorism" is drawn. Wiktionary gives the definition as the following:




The use of mass public communication, usually against a particular individual or group, which






    

  1. incites or inspires acts of terrorism which are statistically probable but happen seemingly at random,
    2. 







    

  1. perpetuates fear through coverage of seemingly random acts of terrorism
    2. 





By that definition, whether or not this is "stochastic terrorism" depends on whether it is statistically probable that protestors loudly protesting in the capitol building will incite or inspire an act of terrorism. Or possibly on whether the sort of communication that caused the protestors to loudly protest is likely to incite or inspire an act of terrorism.


I'd argue that the answer is "no" on both counts though. I think people are far too fast to cry "terrorism" when they actually mean "I don't like it" or "it has chilling effects".