Ben called the _gestalt_ he saw the Blight, after the rogue superintelligence in Vernor Vinge's _A Fire Upon the Deep_: the problem wasn't that people were getting dumber; it was that there was locally coherent coordination away from clarity and truth and towards coalition-building, which was validated by the official narrative in ways that gave it a huge tactical advantage; people were increasingly making decisions that were better explained by their political incentives rather than acting on coherent beliefs about the world—using and construing claims about facts as moves in a power game, albeit sometimes subject to genre constraints under which only true facts were admissible moves in the game.
-When I asked him for specific examples of MIRI or CfAR leaders behaving badly, he gave the example of [MIRI executive director Nate Soares posting that he was "excited to see OpenAI joining the space"](https://intelligence.org/2015/12/11/openai-and-other-news/), despite the fact that [_no one_ who had been following the AI risk discourse](https://slatestarcodex.com/2015/12/17/should-ai-be-open/) [thought that OpenAI as originally announced was a good idea](http://benjaminrosshoffman.com/openai-makes-humanity-less-safe/). Nate had privately clarified that the word "excited" wasn't necessarily meant positively, and in this case meant something more like "terrified."
+When I asked him for specific examples of MIRI or CfAR leaders behaving badly, he gave the example of [MIRI executive director Nate Soares posting that he was "excited to see OpenAI joining the space"](https://intelligence.org/2015/12/11/openai-and-other-news/), despite the fact that [_no one_ who had been following the AI risk discourse](https://slatestarcodex.com/2015/12/17/should-ai-be-open/) [thought that OpenAI as originally announced was a good idea](http://benjaminrosshoffman.com/openai-makes-humanity-less-safe/). Nate [had privately clarified that](https://twitter.com/jessi_cata/status/1462454555925434375) the word "excited" wasn't necessarily meant positively, and in this case meant something more like "terrified."
This seemed to me like the sort of thing where a particularly principled (naïve?) person might say, "That's _lying for political reasons!_ That's _contrary to the moral law!_" and most ordinary grown-ups would say, "Why are you so upset about this? That sort of strategic phrasing in press releases is just how the world works, and things could not possibly be otherwise."
Anna said she didn't want to receive monetary offers from me anymore; previously, she had regarded my custom of throwing money at people to get what I wanted as good-faith libertarianism between consenting adults, but now she was afraid that if she accepted, it would be portrayed in some future Ben Hoffman essay as an instance of her _using_ me. She agreed that someone could have gotten the ideals I had gotten out of "A Sense That More Is Possible", "Raising the Sanity Waterline", _&c._, but there was also evidence from that time pointing the other way (_e.g._, ["Politics Is the Mind-Killer"](https://www.lesswrong.com/posts/9weLK2AJ9JEt2Tt8f/politics-is-the-mind-killer)), that it shouldn't be surprising if people steered clear of controversy.
-I replied: but when forming the original let's-be-apolitical vision in 2008, we did not anticipate that _whether or not I should cut my dick off_ would _become_ a political issue. That was _new evidence_ about whether the original vision was wise! I wasn't trying to do politics with my idiosyncratic special interest; I was trying to _think seriously_ about the most important thing in my life and only do the minimum amount of politics necessary to protect my ability to think. If 2019-era "rationalists" were going to commit a trivial epistemology mistake that interfered with my ability to think seriously about the most important thing in my life, but couldn't correct the mistake, then the 2019-era "rationalists" were _worse than useless_ to me personally. This probably didn't matter causally (I wasn't an AI researcher, therefore I didn't matter), but it might matter timelessly (if I was part of a reference class that includes AI researchers).
+I replied: but when forming the original let's-be-apolitical vision in 2008, we did not anticipate that _whether or not I should cut my dick off_ would _become_ a political issue. That was _new evidence_ about whether the original vision was wise! I wasn't trying to do politics with my idiosyncratic special interest; I was trying to _think seriously_ about the most important thing in my life and only do the minimum amount of politics necessary to protect my ability to think. If 2019-era "rationalists" were going to commit a trivial epistemology mistake that interfered with my ability to think seriously about the most important thing in my life, but couldn't correct the mistake even after it was pointed out, then the 2019-era "rationalists" were _worse than useless_ to me personally. This probably didn't matter causally (I wasn't an AI researcher, therefore I didn't matter), but it might matter timelessly (if I was part of a reference class that includes AI researchers).
Fundamentally, I was skeptical that you _could_ do consisently high-grade reasoning as a group without committing heresy, because of the mechanism that Yudkowsky described in ["Entangled Truths, Contagious Lies"](https://www.lesswrong.com/posts/wyyfFfaRar2jEdeQK/entangled-truths-contagious-lies) and ["Dark Side Epistemology"](https://www.lesswrong.com/posts/XTWkjCJScy2GFAgDt/dark-side-epistemology): the need to lie about lying and cover up cover-ups propagates recursively. Anna in particular was unusually skillful at thinking things without saying them; I thought most people facing similar speech restrictions just get worse at thinking (plausibly[^plausibly] including Yudkowsky), and the problem gets worse as the group effort scales. (It's easier to recommend ["What You Can't Say"](http://www.paulgraham.com/say.html) to your housemates than to put it on a canonical reading list, for obvious reasons.) You _can't_ optimize your group's culture for not-talking-about-atheism without also optimizing against understanding [Occam's razor](https://www.lesswrong.com/posts/f4txACqDWithRi7hs/occam-s-razor); you _can't_ optimize for not questioning gender self-identity without also optimizing against understanding the [37 ways that words can be wrong](https://www.lesswrong.com/posts/FaJaCgqBKphrDzDSj/37-ways-that-words-can-be-wrong).
Despite Math and Wellness Month and my "intent" to take a break from the religious civil war, I kept reading _Less Wrong_ during May 2019, and ended up scoring a couple of victories in the civil war (at some cost to Wellness).
-MIRI researcher Scott Garrabrant wrote a post about how ["Yes Requires the Possibility of No"](https://www.lesswrong.com/posts/G5TwJ9BGxcgh5DsmQ/yes-requires-the-possibility-of-no). Information-theoretically, a signal sent with probability one transmits no information: you can only learn something from hearing a "Yes" if there was some chance that the answer could have been "No". I saw an analogy to my philosophy-of-language thesis, and commented about it: if you want to believe that _x_ belongs to category _C_, you might try redefining _C_ in order to make the question "Is _x_ a _C_?" come out "Yes", but you can only do so at the expense of making _C_ less useful. Meaningful category-membership (Yes) requires the possibility of non-membership (No).
+MIRI researcher Scott Garrabrant wrote a post about how ["Yes Requires the Possibility of No"](https://www.lesswrong.com/posts/G5TwJ9BGxcgh5DsmQ/yes-requires-the-possibility-of-no). Information-theoretically, a signal sent with probability one transmits no information: you can only learn something from hearing a "Yes" if believed that the answer could have been "No". I saw an analogy to my philosophy-of-language thesis, and mentioned it in a comment: if you want to believe that _x_ belongs to category _C_, you might try redefining _C_ in order to make the question "Is _x_ a _C_?" come out "Yes", but you can only do so at the expense of making _C_ less useful. Meaningful category-membership (Yes) requires the possibility of non-membership (No).
-[TODO: explain scuffle on "Yes Requires the Possibility"—
+MIRI research associate Vanessa Kosoy [commented](https://www.lesswrong.com/posts/WwTPSkNwC89g3Afnd/comment-section-from-05-19-2019?commentId=FxSZwECjhgYE7p2du):
- * Vanessa comment on hobbyhorses and feeling attacked
- * my reply about philosophy got politicized, and MDL/atheism analogy
- * Ben vs. Said on political speech and meta-attacks; Goldenberg on feelings
- * 139-comment trainwreck got so bad, the mods manually moved the comments into their own thread https://www.lesswrong.com/posts/WwTPSkNwC89g3Afnd/comment-section-from-05-19-2019
- * based on the karma scores and what was said, this went pretty well for me and I count it as a victory
-
-]
+> I find it unpleasant that you always bring your hobbyhorse in, but in an "abstract" way that doesn't allow discussing the actual object level question. It makes me feel attacked in a way that allows for no legal recourse to defend myself.
+
+I [replied](https://www.lesswrong.com/posts/WwTPSkNwC89g3Afnd/comment-section-from-05-19-2019?commentId=32GPaijsSwX2NSFJi) that that was understandable, but that I hoped it was also understandable that I found it unpleasant that our standard Bayesian philosophy of language somehow got politicized (!?), such that my attempts to do _correct epistemology_ were perceived as attacking people. Imagine living in a world where [posts about the minimum description length principle](https://www.lesswrong.com/posts/mB95aqTSJLNR9YyjH/message-length) were perceived as an attack on Christians—or if that analogy seemed loaded (because our subculture pattern matches atheism as "the good guys"), imagine some racist getting _really interested_ in the statistics of the normal distribution, and posting about the ratio of areas in the right tails of normal distributions with different means. I could see how that would be annoying—maybe even threatening—which would make it all the more satisfying if you could find a _mistake_ in the bastard's math. But if you _couldn't_ find a mistake—if, in fact, the post is on-topic for the forum and correct in the literal things that it literally says, then complaining about the author's motive for being interested in the normal distribution wouldn't seem like an obviously positive contribution to the discourse? I saw the problem, of course, and didn't mean to play dumb about it. But what, realistically, did Kosoy expect the atheist—or the racist, or me—to do?
+
+In a subthread in which I contested Kosoy's characterization of me as a "voice with an agenda which, if implemented, would put [her] in physical danger" ("I don't think of myself as having a lot of strong political beliefs," I said, "but I'm going to take a definite stand here: I am _against_ people being in physical danger"), Ben [pointed out that](https://www.lesswrong.com/posts/WwTPSkNwC89g3Afnd/comment-section-from-05-19-2019?commentId=TXbgr7goFtSAZEvZb)
+
+> Some speech acts lower the message length of proposals to attack some groups, or raise the message length of attempts to prevent such attacks. This is a kind of meta-attack or threat, like concentrating troops on a country's border.
+
+Norms discouraging "political" speech could aggravate the problem, if the response looked "political" but the original threat didn't. If Kosoy wanted to put in the work to explain why my philosophy of language blogging was causing problems for her, she would face legitimate doubt whether her defensive measures would be "admissible".
+
+The trainwreck got so bad that the mods manually [moved the comments to their own post](https://www.lesswrong.com/posts/WwTPSkNwC89g3Afnd/comment-section-from-05-19-2019). Based on the karma scores and what was said, I count it as a "victory" for me.
On 31 May 2019, a [draft of a new _Less Wrong_ FAQ](https://www.lesswrong.com/posts/MqrzczdGhQCRePgqN/feedback-requested-draft-of-a-new-about-welcome-page-for) included a link to "... Not Man for the Categories" as one of Scott Alexander's best essays. I argued that it would be better to cite _almost literally_ any other _Slate Star Codex_ post (most of which, I agreed, were exemplary). I claimed that the following disjunction was true: _either_ Alexander's claim that "There's no rule of rationality saying that [one] shouldn't" "accept an unexpected [X] or two deep inside the conceptual boundaries of what would normally be considered [Y] if it'll save someone's life" was a blatant lie, _or_ one had no grounds to criticize me for calling it a blatant lie, because there's no rule of rationality that says I shouldn't draw the category boundaries of "blatant lie" that way. The mod [was persuaded on reflection](https://www.lesswrong.com/posts/MqrzczdGhQCRePgqN/feedback-requested-draft-of-a-new-about-welcome-page-for?commentId=oBDjhXgY5XtugvtLT), and "... Not Man for the Categories" was not included in the final FAQ. Another "victory."
-But winning "victories" wasn't particularly comforting when I resented this becoming a political slapfight at all. I thought a lot of the objections I faced in the derailed "Possibility of No" thread were insane.
+But winning "victories" wasn't particularly comforting when I resented this becoming a political slapfight at all. I thought a lot of the objections I (and my "allies") faced in the derailed "Possibility of No" thread were insane.
I wrote to Anna and Steven Kaas (who I was trying to "recruit" onto our side of the civil war). In ["What You Can't Say"](http://www.paulgraham.com/say.html), Paul Graham had written, "The problem is, there are so many things you can't say. If you said them all you'd have no time left for your real work." But surely that depends on what _is_ one's real work. For someone like Paul Graham, whose goal was to make a lot of money writing software, "Don't say it" (except for this one meta-level essay) was probably the right choice. But someone whose goal is to improve our collective ability to reason, should probably be doing _more_ fighting than Paul Graham (although still preferably on the meta- rather than object-level), because political restrictions on speech and thought directly hurt the mission of "improving our collective ability to reason", in a way that they don't hurt the mission of "make a lot of money writing software."
-----
-While visiting Valinor on 7 August 2019, the older son (age 2¾ years) asked me, "Why are you a boy?"
+While visiting Valinor on 7 August 2019, Merlin Blume (age 2¾ years) asked me, "Why are you a boy?"
After a long pause, I said, "Yes," as if I had misheard the question as "Are you a boy?" I think it was a motivated mishearing: it was only after I answered that I consciously realized that's not what the kid asked.
(It does not, actually, have a happy ending where everyone comes to their senses.)
+------
+
+While visiting Valinor on 4 February 2020, I remember my nose dripping while I was holding Koios, the baby. Alicorn offered me a tissue. I asked if I shouldn't be holding the baby while my nose was dripping and I therefore plausibly had a cold. She said it was fine. On the topic of possible sickness, I said that I hoped the novel coronavirus people were talking about didn't go pandemic.
+
+It did. The Berkeley rats took social distancing guidelines very seriously, so it would be a while before I could visit again.
+
+------
+
+On 10 February 2020, Scott Alexander published ["Autogenderphilia Is Common and Not Especially Related to Transgender"](https://slatestarcodex.com/2020/02/10/autogenderphilia-is-common-and-not-especially-related-to-transgender/), an analysis of the results of the autogynephilia/autoandrophilia questions on the recent _Slate Star Codex_ survey.
+
+I appreciated the gesture of getting real data, but I was deeply unimpressed with Alexander's analysis for reasons that I found difficult to write up in a timely manner. Three years later, I eventually got around to [polishing my draft and throwing it up as a standalone post](/2023/Feb/reply-to-scott-alexander-on-autogenderphilia/), rather than cluttering the present narrative with my explanation.
+
+Briefly, based on eyballing the survey data, Alexander proposes "if you identify as a gender, and you're attracted to that gender, it's a natural leap to be attracted to yourself being that gender" as a "very boring" theory, but on my worldview, a hypothesis that puts "gay people (cis and trans)" in the antecedent is _not_ boring and actually takes on a big complexity penalty: I just don't think the group of gay men _and_ lesbians _and_ straight males with female gender identities _and_ straight females with male gender identities have much in common with each other, except sociologically (being "queer"), and by being human.
+
+(I do like the [hypernym](https://en.wikipedia.org/wiki/Hyponymy_and_hypernymy) _autogenderphilia_.)
+
-------
There's another extremely important part of the story that _would_ fit around here chronologically, but I again find myself constrained by privacy norms: everyone's common sense of decency (this time, even including my own) screams that it's not my story to tell.
-----
-[TODO: pandemic starts]
-
-[TODO: "Autogenderphilia Is Common" https://slatestarcodex.com/2020/02/10/autogenderphilia-is-common-and-not-especially-related-to-transgender/]
-
------
-
On 1 June 2020, I received a Twitter DM from _New York Times_ reporter Cade Metz, who said he was "exploring a story about the intersection of the rationality community and Silicon Valley". I sent him an email saying that I would be happy to talk, but that I'd actually been pretty disappointed with the community lately: I was worried that the social pressures of trying to _be_ a "community" and protect the group's status (_e.g._, from _New York Times_ reporters who might portray us in an unflattering light??) incentivize people to compromise on the ideals of _systematically correct reasoning_ that made the community valuable in the first place.
He never got back to me. Three weeks later, all existing _Slate Star Codex_ posts were taken down.
How could I make this rigorous? Did I want to be talking about the _variance_ of my features conditional on category-membership? Was "connectedness" intrinsically the what I wanted, or was connectedness only important because it cut down the number of possibilities? (There are 8!/(6!2!) = 28 ways to choose two elements from `{1..8}`, but only 7 ways to choose two contiguous elements.) I thought connectedness _was_ intrinsically important, because we didn't just want _few_ things, we wanted things that are _similar enough to make similar decisions about_.
-I put the question to a few friends (Subject: "rubber duck philosophy"), and Jessica said that my identification of the variance as the key quantity sounded right: it amounted to the expected squared error of someone trying to guess the values of the features given the category. It was okay that this wasn't a purely information-theoretic criterion, because for problems involving guessing a numeric quantity, bits that get you closer to the right answer were more valuable than bits that didn't.
+I put the question to a few friends in July 2020 (Subject: "rubber duck philosophy"), and Jessica said that my identification of the variance as the key quantity sounded right: it amounted to the expected squared error of someone trying to guess the values of the features given the category. It was okay that this wasn't a purely information-theoretic criterion, because for problems involving guessing a numeric quantity, bits that get you closer to the right answer were more valuable than bits that didn't.
------
projects / Ultimately_Untrue_Thought.git / blobdiff