Astral Codex Ten

Charles Lehman

11 min read

Charles Lehman

Article

Charles Lehman is a recurring person in the Astral Codex Ten archive, appearing 3 times across 3 issues between July 24, 2024 and April 15, 2025. The archive places it in contexts such as “Charles Lehman responds to my posts on mentally ill homeless people”; “Later, Charles Lehman wrote a response describing his plan and arguing why it’s necessary”; “Charles Lehman writes (X) :“. It most often appears alongside America, Biden, China.

Metadata

  • Category: People
  • Mention count: 3
  • Issue count: 3
  • First seen: July 24, 2024
  • Last seen: April 15, 2025

Appears In

Source Context

Recovered passages from the original issue text. When the raw archive preserved outbound links inside the source passage, they are listed directly under the quote.

Links for July 2024
July 24, 2024 · Original source
37: Charles Lehman responds to my posts on mentally ill homeless people: Serious Mental Illness Is An Optimization Problem.
Inline links: Serious Mental Illness Is An Optimization Problem
Links For September 2024
September 12, 2024 · Original source
12: A while back I wrote a piece saying people needed to be clearer about what their “GET TOUGH” plans for dealing with mentally ill homeless people really meant. Later, Charles Lehman wrote a response describing his plan and arguing why it’s necessary. Most recently, Ozy has written a response to Charles, basically expressing fear that Charles’ plan will unnecessarily commit a bunch of harmless well-functioning people. I bet Charles’ response will be that no, this isn’t what he wants at all, to which my response will be that this is why you need to be clearer about what you mean. That is, I’m sure Charles wants to only commit people who need commitment, and not commit people who don’t, but he hasn’t explained the mechanism by which a fallible court system and medical system will ensure that this actually happens, and those are the kinds of details that I’m most interested in.
Inline links: I wrote a piece saying people needed to be clearer, wrote a response, Ozy has written a response to Charles
Highlights From The Comments On POSIWID
April 15, 2025 · Original source
Balanced: Systems can fail for many reasons. Sometimes it’s just a hard problem with tradeoffs. Sometimes it’s been perverted from its original goal by special interests. Sometimes it’s some third thing. Usually it’s a combination of all of these. You can’t know for sure until you look at it closely. …then I think the people who use the phrase want to imagine that they’re pushing people from Naive to Balanced. But I think the last person to hold the Naive perspective died sometime in the 1980s, and in real life POSIWID is mostly used to push people from the Balanced to the Paranoid perspective without actually looking at the system involved or arguing the case. And second, the explanation above is just the most popular of about a half-dozen different exegeses that commenters offered. So if you do want to communicate the thing I suggested above - which, reminder, is: If a system consistently fails at its stated purpose, but people don’t change it, consider that the stated purpose is less important than some actual, hidden purpose, at which it is succeeding …I think you should just say that, instead of the confusing version that half of your audience will misinterpret, and which incorrectly implies that this always happens. When people insist on the confusing and inappropriately-strong version, I start to suspect that the confusingness is a feature, letting them smuggle in connotations that people would otherwise correctly challenge. Hopefully this will become clearer as I answer your comments one by one, starting with: Charles Lehman writes (X): The actually useful insight from POSIWID is the negative corollary: there is "no point in claiming that the purpose of a system is to do what it constantly fails to do." This was the original, less-catchy form, but it seems exactly the same to me, and equally wrong. For example, Iran’s intelligence agency consistently fails to prevent Israel from infiltrating and attacking their nuclear program. But it’s very useful to claim that their purpose is to prevent this! If we both try to predict the behavior of Iranian intelligence, and I’m allowed to use the hypothesis “their purpose is preventing Israeli infiltration”, and you’re not allowed to use that hypothesis, I will consistently outpredict you. For example, I’ll be expecting them to interview security staff to see which ones are Israeli spies, try to intercept Israeli communications, and do other espionage activities, and you’ll still be stuck wondering whether they might take up gardening or ballet dancing. The clear, natural language expression of this useful hypothesis is “the purpose of Iran’s intelligence forces is to prevent Israeli infiltration, but they usually fail”. This clear, natural language expression is great and tells you everything that you need to know. POSIWID (or its inverse) adds nothing except an attempt to ban this excellent and communicative expressive technology, in favor of some other vague meaning of “purpose” which can’t hang together and which nobody can really explain. Ersatz writes: “I thought the meaning was more something like “the system took these side effects into account and still considered that what it was doing was net positive in expectation, so the side effects are as much part of the system's purpose as the ‘positive’ outcomes”. This is just diluting the word “purpose” into incoherence. I think it’s useful linguistic technology to be able to say “The New York bus system both transports people from place to place, and emits lots of carbon dioxide. Its purpose is the transportation, and the carbon dioxide is an unfortunate side effect”. If the goal of POSIWID is to insist that no, the transportation and the CO2 emissions are equally purposeful, or that we’re not allowed to talk about that question, then it sabotages our ability to communicate clearly, for no apparent gain. Consider the New York carbon dioxide system, a hypothetical government department dedicated to emitting as much carbon dioxide as possible (why? to own the libs, of course!) Sometimes people drive motorcycles along its vast network of CO2 pipes, allowing them to travel from place to place. Is this exactly the same as the New York bus system? No? Why not? A natural answer would be “Because the bus system is aimed at transportation, but emits some CO2 as a minor side effect, and the CO2 system is aimed at emission, but facilitates transport as a minor side effect.” If your objection is going to be that instead of considering purpose, you should restrict yourself to saying that one transports more people than it emits CO2, and the other emits more CO2 than it helps transport people, you’ve now legislated that we must list exactly how many people the New York bus system transports, and how much carbon dioxide it emits, and how many ants it crushes, etc, every time we talk about it. But most people who talk about the bus system don’t know these statistics, and don’t have to. It’s sufficient (and linguistically convenient) to say “Its purpose is transportation, not CO2 emission or ant-crushing. Andrew Pearson writes: The steelman of the phrase, which might be what Stafford Beer had in mind when he coined the phrase, is that in large complex organisations it's very hard for individual workers to see a link between what they do and the stated purpose of the organisation - people just get on and do what they're told, and orient themselves more towards "doing what they're already doing, just more effectively" than towards "fulfilling the stated purpose of the organisation". I don’t think this is a great steelman. True, the average employee’s actions don’t obviously connect to the organization’s stated purpose. But the average employee’s actions also don’t obviously connect to what the organization does. If the stated purpose of the US military is to protect America, and what it does is bomb Middle Eastern weddings, both of those are equally remote from the day-to-day life of some low-level employee who just counts the number of screws in a warehouse. Aashish Reddy writes: I think POSIWID is best applied to bureaucracies or large structures where the reason bad outcomes occur is not because of difficult battles with reality (like government or hospitals or the Ukrainian military), but because of the way incentives are set up in the system. If someone said, “the purpose of the Civil Service is to drive through new, innovative ways of delivering rapid change!”, that would clearly be absurd. That may be their goal, or how they see themselves; but the purpose of the system is not defined by either of those things. If it was, they wouldn’t incentivise caution and slowness. Whether that’s good or not, the purpose of the Civil Service is best approximated by what it does! I want to pay more attention to the word “goal” in the second sentence of the second paragraph: “[Driving innovative change] may be their goal, or how they see themselves.” It sounds like Aashish think it’s useful to use the word “goal” to discuss what a system is trying to do, separately from what it does or doesn’t accomplish. I agree! I just think “purpose” is a synonym for goal. If you use POSIWID, you have to posit some kind of weird new ontology where “purpose” means the opposite of “goal”. If you don’t use POSIWID, you can just keep the words “purpose” and “goal” having their regular everyday meaning, and describe this state of affairs with phrases like “The goal/purpose of the Civil Service is to deliver rapid change, but due to perverse incentives, its actual effect is to prevent change.” Kay writes: For example, the US policing and criminal justice system and prisons seem to continually over imprisons people in general, and especially people of colour. These systems seem not to be changed, while they ostensibly truly could be. So to me this seems that the system may be working "to purpose" for the actors who want it to work that way. This is interesting because Kay is using a left-wing example: “The criminal justice system imprisons too many people of color, so its purpose is to oppress black people”. But one of the tweets in my original post was close to its right-wing opposite: “The criminal justice system consistently lets criminals off with a slap on the wrist, so its purpose is to get people raped and murdered.” As long as people are thinking in these terms, they’re going to be prey for whichever conspiracy theory best suits their pre-existing prejudices. I think both of these are worse than the Balanced View version: Some people care a lot about keeping people safe from crime, but other people care a lot about the human rights of suspects and convicts. The incarceration rate is a balance between these two forces, with some lesser contribution from sinister forces like private prison owners who want to increase incarceration to line their pockets. This has the advantage of being obviously true (there are pro-tough-on-crime activist groups and pro-soft-on-crime activist groups, they both do effective activism for their chosen cause, and the exact incarceration rate depends on which ones are in the ascendant and have the ear of politicians), and not being a conspiracy theory that forces you to believe that the government is a monolithic entity that really wants people to be raped and murdered. NegatingSilence writes: Come on, obviously "The Purpose of a System is What it Does" is meant to draw your attention to the incentives in cases where something different is happening than is supposed to be happening. My government purposefully raised housing prices for the benefit of people who own assets. They talk about "trying" to make things more affordable, but somehow they only succeeded in raising the cost by 500% in nominal terms. What a curious result. I don’t know what country this person is in. But in my country, housing prices are high because of a combination of all of the following: Citizens want to preserve “neighborhood character”: they currently live in a low-density low-crime low-traffic pretty suburb, they want it to remain a low-density low-crime low-traffic pretty suburb, and they worry that building new homes threatens that status.
Inline links: writes (X), writes, writes, writes, writes