—which would have been the end of the story, _except that_, as I explained in a subsequent–subsequent post, ["A Hill of Validity in Defense of Meaning"](/2023/Jul/a-hill-of-validity-in-defense-of-meaning/), in late 2018, Eliezer Yudkowsky prevaricated about his own philosophy of language in the service of the gender-identity coalition, and my unsuccessful attempts to get him to clarify led me and allies to conclude that Yudkowsky and his "rationalists" were corrupt.
-Anyway, given that the "rationalists" were fake and that we needed something better, there remained the question of what to do about that, and how to relate to the old thing, and the operators of the marketing machine for the old thing.
+Anyway, given that the "rationalists" were fake and that we needed something better, there remained the question of what to do about that, and how to relate to the old thing.
-_I_ had been hyperfocused on prosecuting my Category War, but the reason Michael and Ben and Jessica were willing to help me out on that was not because they particularly cared about the gender and categories example, but because it seemed like a manifestation of a _more general_ problem of epistemic rot in "the community".
+_I_ had been hyperfocused on prosecuting my Category War, but the reason Michael Vassar and Ben Hoffman and Jessica Taylor were willing to help me out on that was not because they particularly cared about the gender and categories example, but because it seemed like a manifestation of a more general problem of epistemic rot in "the community".
Ben had [previously](http://benjaminrosshoffman.com/givewell-and-partial-funding/) [written](http://benjaminrosshoffman.com/effective-altruism-is-self-recommending/) a lot [about](http://benjaminrosshoffman.com/openai-makes-humanity-less-safe/) [problems](http://benjaminrosshoffman.com/against-responsibility/) [with](http://benjaminrosshoffman.com/against-neglectedness/) Effective Altruism. Jessica had had a bad time at MIRI, as she had told me back in March, and would [later](https://www.lesswrong.com/posts/KnQs55tjxWopCzKsk/the-ai-timelines-scam) [write](https://www.lesswrong.com/posts/MnFqyPLqbiKL8nSR7/my-experience-at-and-around-miri-and-cfar-inspired-by-zoe) [about](https://www.lesswrong.com/posts/pQGFeKvjydztpgnsY/occupational-infohazards). To what extent were my thing, and Ben's thing, and Jessica's thing, manifestations of "the same" underlying problem? Or had we all become disaffected with the mainstream "rationalists" for our own idiosyncratic reasons, and merely randomly fallen into each other's, and Michael's, orbit?
From this, Jessica derived the moral that when people are doing something that seems obviously terrible and in bad faith, it could help to publicly explain why the abstract thing is bad, without accusing anyone. This made sense because people didn't want to be held to standards that other people aren't being held to: a call-out directed at oneself personally could be selective enforcement, but a call-out of the abstract pattern invited changing one's behavior if the new equilibrium looked better.
-Michael said that part of the reason this worked was because it represented a clear threat to skapegoat, while also _not_ skapegoating, and not surrendering the option to do so later; it was significant that Jessica's choice of example positioned her on the side of the powerful social-justice coalition.
+Michael said that part of the reason this worked was because it represented a clear threat to scapegoat, while also _not_ scapegoating, and not surrendering the option to do so later; it was significant that Jessica's choice of example positioned her on the side of the powerful social-justice coalition.
------
I had an interesting interaction with Somni, one of the "Meeker Four"—presumably out on bail at this time?—on 12 December 2019.
-I told her, from a certain perspective, it's surprising that you spend so much time complaining about CfAR, Anna Salamon, Kelsey Piper, _&c._, but _I_ seemed to get along fine with her—because "naïvely", one would think that my views were _so much worse_. Was I getting a pity pass because she thought false consciousness was causing me to act against my own transfem class interests? Or what?
+I told her, from a certain perspective, it's surprising that you spend so much time complaining about CfAR, Anna Salamon, Kelsey Piper, _&c._, but _I_ seemed to get along fine with her—because "naïvely", one would think that my views were so much worse. Was I getting a pity pass because she thought false consciousness was causing me to act against my own transfem class interests? Or what?
-To be absolutely clear about my terrible views, I said that I was privately modeling a lot of transmisogyny complaints as something like—a certain neurotype-cluster of non-dominant male is latching onto locally-ascendant social-justice ideology in which claims to victimhood can be leveraged into claims to power. Traditionally, men are moral agents, but not patients; women are moral patients, but not agents. If weird non-dominant men aren't respected if identified as such (because low-ranking males aren't valuable allies, and don't have intrinsic moral patiency of women), but _can_ get victimhood/moral-patiency points for identifying as oppressed transfems, that creates an incentive gradient for them to do so, and no one was allowed to notice this except me, because everyone prefers to stay on the good side of social-justice ideology unless they have Something to Protect that requires defying it.
+In order to be absolutely clear about my terrible views, I said that I was privately modeling a lot of transmisogyny complaints as something like—a certain neurotype-cluster of non-dominant male is latching onto locally-ascendant social-justice ideology in which claims to victimhood can be leveraged into claims to power. Traditionally, men are moral agents, but not patients; women are moral patients, but not agents. If weird non-dominant men aren't respected if identified as such (because low-ranking males aren't valuable allies, and don't have intrinsic moral patiency of women), but _can_ get victimhood/moral-patiency points for identifying as oppressed transfems, that creates an incentive gradient for them to do so, and no one was allowed to notice this except me, because everyone prefers to stay on the good side of social-justice ideology unless they have Something to Protect that requires defying it.
Somni said that it was because I was being victimized by the same forces of gaslighting, and that I wasn't lying about my agenda. Maybe she _should_ be complaining about me?—but I seemed to be following a somewhat earnest epistemic process, whereas Kelsey, Scott, and Anna were not. If I were to start going, "Here's my rationality org; rule #1: no transfems (except me); rule #2, no telling people about rule #1", then she would talk about it.
-I would later remark to Anna (Subject: "a peek behind the fog of war; and, [redacted]", 6 February 2020) that Somni and Ziz saw themselves as being oppressed by people's hypocritical and manipulative social perceptions and behavior.
-
-Merely using the appropriate language ("Somni ... she", _&c._) protected her against threats from the Political Correctness police, but it actually didn't protect against threats from _them_. It was as if the mere fact that I wasn't optimizing for PR (lying about my agenda, as Somni said) was what made me not a direct enemy (although still a collaborator) in their eyes.
+I would later remark to Anna that Somni and Ziz saw themselves as being oppressed by people's hypocritical and manipulative social perceptions and behavior. Merely using the appropriate language ("Somni ... she", _&c._) protected her against threats from the Political Correctness police, but it actually didn't protect against threats from _them_. It was as if the mere fact that I wasn't optimizing for PR (lying about my agenda, as Somni said) was what made me not a direct enemy (although still a collaborator) in their eyes.
--------
Scott said he might agree with me if he thought the world-model-clarity _vs._ utilitarian benefit tradeoff was unfavorable—or if he thought it had the chance of snowballing like in his "Kolmogorov Complicity and the Parable of Lighting".
-... I pointed out that what sex people are is more relevant to human social life than whether lightning comes before thunder. He said that the problem in his parable was that people were being made ignorant of things, whereas in the transgender case, no one was being kept ignorant; their thoughts were just following a longer path.
+I pointed out that what sex people are is more relevant to human social life than whether lightning comes before thunder. He said that the problem in his parable was that people were being made ignorant of things, whereas in the transgender case, no one was being kept ignorant; their thoughts were just following a longer path.
I had technical reasons to be very skeptical of the claim that no one was "really" being kept ignorant. If you're sufficiently clever and careful and you remember how language worked when Airstrip One was still Britain, then you can still think, internally, and express yourself as best you can in Newspeak. But a culture in which Newspeak is mandatory, and all of Oceania's best philosophers have clever arguments for why Newspeak doesn't distort people's beliefs ... doesn't seem like a nice place to live, right? Doesn't seem like a culture that can solve AI alignment, right?
If you have the visual metaphor, but you don't have the math, you might think that there's nothing intrinsically wrong with squiggly or discontinuous category "boundaries", just as there's nothing intrinsically wrong with Alaska not being part of the contiguous U.S. states. It may be inconvenient that you can't drive from Alaska to Washington without going through Canada, and we have to deal with the consequences of that, but there's no sense in which it's _wrong_ that the borders are drawn that way: Alaska really is governed by the United States.
-But if you _do_ have the math, a moment of introspection will convince you that the analogy between category "boundaries" and national borders is not a particularly deep or informative one.
+But if you _do_ have the math, a moment of introspection will convince you that the analogy between category "boundaries" and national borders is shallow.
A two-dimensional political map tells you which areas of the Earth's surface are under the jurisdiction of what government. In contrast, category "boundaries" tell you which regions of very high-dimensional configuration space correspond to a word/concept, which is useful _because_ that structure is useful for making probabilistic inferences: you can use your observastions of some aspects of an entity (some of the coordinates of a point in configuration space) to infer category-membership, and then use category membership to make predictions about aspects that you haven't yet observed.
But the trick only works to the extent that the category is a regular, non-squiggly region of configuration space: if you know that egg-shaped objects tend to be blue, and you see a black-and-white photo of an egg-shaped object, you can get _close_ to picking out its color on a color wheel. But if egg-shaped objects tend to blue _or_ green _or_ red _or_ gray, you wouldn't know where to point to on the color wheel.
-The analogous algorithm applied to national borders on a political map would be to observe the longitude of a place, use that to guess what country the place is in, and then use the country to guess the latitude—which isn't typically what people _do_ with maps. Category "boundaries" and national borders might both be _illustrated_ similarly in a two-dimensional diagram, but philosophically, they're very different entities. The fact that Scott Alexander was appealing to national borders to explain why gerrymandered categories were allegedly okay, showed that he Didn't Get It.
+The analogous algorithm applied to national borders on a political map would be to observe the longitude of a place, use that to guess what country the place is in, and then use the country to guess the latitude—which isn't typically what people do with maps. Category "boundaries" and national borders might both be _illustrated_ similarly in a two-dimensional diagram, but philosophically, they're different entities. The fact that Scott Alexander was appealing to national borders to explain why gerrymandered categories were allegedly okay, showed that he Didn't Get It.
-I still had some deeper philosophical problems to resolve, though. If squiggly categories were less useful for inference, why would someone _want_ a squiggly category boundary? Someone who said, "Ah, but I assign _higher utility_ to doing it this way", had to be messing with you. Where would such a utility function come from? Intuitively, it had to be precisely _because_ squiggly boundaries were less useful for inference; the only reason you would _realistically_ want to do that would be to commit fraud, to pass off pyrite as gold by redefining the word "gold."
+I still had some deeper philosophical problems to resolve, though. If squiggly categories were less useful for inference, why would someone _want_ a squiggly category boundary? Someone who said, "Ah, but I assign higher utility to doing it this way", had to be messing with you. Where would such a utility function come from? Intuitively, it had to be precisely _because_ squiggly boundaries were less useful for inference; the only reason you would realistically want to do that would be to commit fraud, to pass off pyrite as gold by redefining the word "gold."
That was my intuition. To formalize it, I wanted some sensible numerical quantity that would be maximized by using "nice" categories and get trashed by gerrymandering. [Mutual information](https://en.wikipedia.org/wiki/Mutual_information) was the obvious first guess, but that wasn't it, because mutual information lacks a "topology", a notion of _closeness_ that made some false predictions better than others by virtue of being "close".
------
-[TODO:
- * Yudkowsky made a stray remark about social media causing people to say crazy thing
- * I got enraged, posted a couple Tweets, including a preview of "Unnatural Categories"
+I decided on "Unnatural Categories Are Optimized for Deception" as the title for my advanced categorization thesis. Writing it up was a large undertaking. There were a lot of nuances to address and potential objections to preëmpt, and I felt that I had to cover everything. (A reasonable person who wanted to understand the main ideas wouldn't need so much detail, but I wasn't up against reasonable people who wanted to understand.)
-https://twitter.com/ESYudkowsky/status/1304824253015945216
-> Hypothesis: social media has an effect promoting Terrible Straw Arguments to being used by many actual people. One crazy on Side A makes a bad argument. Side B subtweets with a refutation and that gets a million views. So people on Side A hear about it as Side A's argument.
->
-> And the result looks so incredibly surreal. People be going "2 + 3 = 4" and the replies are things like "lol but 4 is a prime number you moron". The entire conversation has been implicitly selected to sound insane to the other side.
+In September 2020, Yudkowsky Tweeted [something about social media incentives prompting people to make insane arguments](https://twitter.com/ESYudkowsky/status/1304824253015945216).
+Something in me boiled over. The Tweet was fine in isolation, but I rankled at it in the context of his own incentive-driven insanity remaining unaddressed. I left [a snarky reply](https://twitter.com/zackmdavis/status/1304838486810193921) and [vented on my timeline](https://twitter.com/zackmdavis/status/1304838346695348224) (with preview images from the draft of "Unnatural Categories Are Optimized for Deception"):
> Who would have thought getting @ESYudkowsky's robot cult to stop trying to trick me into cutting my dick off (independently of the empirical facts determining whether or not I should cut my dick off) would involve so much math?? OK, I guess the math part isn't surprising, but—
-https://twitter.com/zackmdavis/status/1304838346695348224
-
-
-https://twitter.com/zackmdavis/status/1304838486810193921
-> Or, "You're not standing in defense of truth if you insist on a symbol like '2' or '+', brought explicitly into question, being used with some particular meaning." (I'm still very angry at you! This shouldn't be hard to clear up in public!) 1/2
->
-> If not the object-level application on Twitter, at least the abstract philosophy-of-language issue on Less Wrong?! Please!?! I've been working on a more detailed explanation that should be ready within a few weeks. 2/2
-
- * something in my boiled over, and I eventually ended up staying up late writing an angry email
- * this ruins my chances for being a "neutral" bridge between the Vassarites and the Caliphate, but that's OK
- * mostly reproduce below (with a few redactions for either brevity or compliance with privacy norms, but I'm not going to clarify which)
-]
-[TODO: "out of patience" email]
+My rage-boil continued into staying up late writing him an angry email, which I mostly reproduce below (with a few redactions for either brevity or compliance with privacy norms, but I'm not going to clarify which).
> To: Eliezer Yudkowsky <[redacted]>
> Cc: Anna Salamon <[redacted]>
> Date: Sunday 13 September 2020 11:02 _a.m._
> Subject: Re: out of patience
>
-> [... redacted] The sinful and corrupted part wasn't the _initial_ Tweets; the sinful and corrupted part is this **bullshit stonewalling** when your Twitter followers and me and Michael and Ben and Sarah and [redacted] and Jessica tried to point out the problem. I've _never_ been arguing against your private universe [... redacted]; the thing I'm arguing against in ["Where to Draw the Boundaries?"](https://www.lesswrong.com/posts/esRZaPXSHgWzyB2NL/where-to-draw-the-boundaries) (and **my [unfinished draft sequel](https://github.com/zackmdavis/Category_War/blob/cefa98c3abe/unnatural_categories_are_optimized_for_deception.md)**, although that's more focused on what Scott wrote) is the **_actual text_ you _actually published_, not your private universe.**
+> [... redacted] The sinful and corrupted part wasn't the _initial_ Tweets; the sinful and corrupted part is this **bullshit stonewalling** when your Twitter followers and me and Michael and Ben and Sarah and ["Riley"] and Jessica tried to point out the problem. I've _never_ been arguing against your private universe [... redacted]; the thing I'm arguing against in ["Where to Draw the Boundaries?"](https://www.lesswrong.com/posts/esRZaPXSHgWzyB2NL/where-to-draw-the-boundaries) (and **my [unfinished draft sequel](https://github.com/zackmdavis/Category_War/blob/cefa98c3abe/unnatural_categories_are_optimized_for_deception.md)**, although that's more focused on what Scott wrote) is the **_actual text_ you _actually published_, not your private universe.**
>
> [... redacted] you could just **publicly clarify your position on the philosophy of language** the way an intellectually-honest person would do if they wanted their followers to have correct beliefs about the philosophy of language?!
>
[TODO: Sep 2020 categories clarification from EY—victory?!
https://www.facebook.com/yudkowsky/posts/10158853851009228
_ex cathedra_ statement that gender categories are not an exception to the rule, only 1 year and 8 months after asking for it
+ * this ruins my chances for being a "neutral" bridge between the Vassarites and the Caliphate, but that's OK
]
-And really, that _should_ have been the end of the story. At the trifling cost of two years of my life, we finally got a clarification from Yudkowsky that you can't define the word _woman_ any way you like. I didn't think I was entitled to anything more than that. I was satisfied. If I hadn't been further provoked, I wouldn't have occasion to continue waging the robot-cult religious civil war.
-
-----
If this were an autobiography (which existed to tell my life story) rather than a topic-focused memoir (which exists because my life happens to contain this Whole Dumb Story which bears on matters of broader interest, even if my life would not otherwise be interesting), there's a dramatic episode that would fit here chronologically.
-----
-I still published ["Unnatural Categories Are Optimized for Deception"](https://www.lesswrong.com/posts/onwgTH6n8wxRSo2BJ/unnatural-categories-are-optimized-for-deception) in January 2021.
+I published ["Unnatural Categories Are Optimized for Deception"](https://www.lesswrong.com/posts/onwgTH6n8wxRSo2BJ/unnatural-categories-are-optimized-for-deception) in January 2021.
I wrote back to Abram Demski regarding his comments from fourteen months before: on further thought, he was right. Even granting my point that evolution didn't figure out how to track probability and utility separately, as Abram had pointed out, the _fact_ that it didn't meant that not tracking it could be an effective AI design. Just because evolution takes shortcuts that human engineers wouldn't didn't mean shortcuts are "wrong". (Rather, there are laws governing which kinds of shortcuts _work_.)
Abram was also right that it would be weird if reflective coherence was somehow impossible: the AI shouldn't have to fundamentally reason differently about "rewriting code in some 'external' program" and "rewriting 'its own' code." In that light, it made sense to regard "have accurate beliefs" as _merely_ a convergent instrumental subgoal, rather than what rationality is about—as sacrilegious as that felt to type.
And yet, somehow, "have accurate beliefs" seemed _more fundamental_ than other convergent instrumental subgoals like "seek power and resources". Could this be made precise? As a stab in the dark, was it possible that the [theorems on the ubiquity of power-seeking](https://www.lesswrong.com/posts/6DuJxY8X45Sco4bS2/seeking-power-is-often-robustly-instrumental-in-mdps) might generalize to a similar conclusion about "accuracy-seeking"? If it _didn't_, the reason why it didn't might explain why accuracy seems more fundamental.
+
+------
+
+And really, that _should_ have been the end of the story. At the trifling cost of two years of my life, we finally got a clarification from Yudkowsky that you can't define the word _woman_ any way you like. I didn't think I was entitled to anything more than that. I was satisfied. If I hadn't been further provoked, I wouldn't have occasion to continue waging the robot-cult religious civil war.
+
+(To be continued.)
projects / Ultimately_Untrue_Thought.git / blobdiff