#org_design

1. The Analects

Earlier this year, I purchased one copy each of The Analects of Confucius and the Tao Te Ching (or Dao Dejing). I had long intended to read more philosophy (and felt guilty about not doing so). The design of each book was gorgeous, and the contents were short.

What struck me the most about The Analects was the similarity in the problem of governance between ancient China and contemporary America, despite the astounding gulf in energy, knowledge, and democratic governance.

Over the past decade, I have developed a better ability to read people, to understand their positioning, nature, and motivations. Based on my reading of the Analects (with the help of the translator's notes in my copy), it is my opinion that Confucius was a very specific person.

He was fascinated with knowledge and learning. Likewise, he was fascinated not only with government, but with the art of governing, which pervades everyday life. He loved order, yes, but not only could he separate manners from substance, he was also kind.

He is attempting to promote the development of the youth, organize the hierarchy of the state, and promote a good quality of governance. From his body of thought, we can deduce that his objective is a benevolent government obtained through a combination of strength and humility. Why does he want this? The man lived into his seventies, and eschews resentment. It would seem that he wants it for the benefit that it provides to others.

These days, I often speak of bits and limits on information. As I told arcticdementor...

Nearly 7,500 words is enough already! A 117,000-word post is called a "book."

Language divides the world into categories. The more finely-grained the categories, the more language you need. It's like a digital picture – if you wish to show every beautiful whisker on your new kitten's little head, then you need a lot of pixels, while if you are content to show him as a vague, cat-shaped blur, then you don't need very many.

The version of the Analects on Project Gutenberg (though it is not the one I used) is 30,000 words. It is divided into 20 books, yielding an average of 1,500 words each. As an example, book 7 is 1,590 words and is divided in 37 chapters, yielding an average of about 43 words each. This is about the length of a contemporary 280-character tweet.

In a 280-character tweet, how much room is there for qualifiers and constraints? How much room is there to explain the substantive basis for a rule, rather than the rule itself? Confucius does provide different answers to different people, but he is also trying to guide people in a particular direction. There is little room for him to say, "You already study enough. Here, for today, why don't you drink and be merry?"

While a philosopher still lives, he can do the labor of adapting his doctrine to each individual, or of altering it according to the changing circumstances of society. The real Confucius was a living, breathing, feeling, complicated man, and his written doctrine is necessarily much simpler than everything he was. It is a projection of the rich and substantive, high-dimensionality Confucius onto the thin, insubstantial, and low-dimensionality text of the Analects, like the shadow of a man against a paper screen.

Confucius is a human being, while the Analects is capital.

If Confucius were alive today, he and I might be friends, or at least discourse pals.

2. Mitigated Chaos

Confucius describes himself as someone who transmits information, rather than innovating. From my perspective, the function of the blog Mitigated Chaos is to explain things that are obvious, in clear language.

Perhaps it is merely mental exercise. I am a double-contrarian. I invert inversion and deconstruct deconstruction. This is more difficult than the original maneuver, and demonstrating it implies a superior place in the intellectual hierarchy (or so one imagines). I certainly like to imagine that I am acting out of good motives, defending the weak and promoting prosperity over the long term, but perhaps I have tricked myself – the latter view might be the belief of my most perceptive left-leaning critics.

Nonetheless, in the past year it has been confirmed, both on Tumblr and Twitter, that Mitigated Chaos is very dense.

It isn't just that LLMs light up when reading a post from Mitigated Chaos. They drop words like "analytical," "articulate," "thorough," "eloquent," "insightful," and "sophisticated." Rather, by experiment, it takes roughly five times as many words for someone outside my circle to understand what I have written as it does for someone inside my circle. That's not an estimate. I've measured it.

As the author of Mitigated Chaos, with the help of some modest translator's notes, much of what Confucius says in the Analects seems obvious to me. I summarized it, very briefly, as, "States run well when men do their jobs; states don't run well when men don't do their jobs." (This doesn't mean that I think everything he said is correct, just that it's easy to understand what his likely reasoning is.)

Perhaps others would find the same translation that I read to be more challenging reading.

The most important thing I learned from reading The Analects was not any specific thing that Confucius said. The most important thing I learned was that an ancient Chinese philosopher said these things 2,500 years ago.

Where Confucius and I agree, even on the meta level, this is not, "just some idea made up by a nobody on the Internet," but, right or wrong, an ancient observation on human nature that has endured throughout the ages, not limited to one particular culture or time.

3. The Dao

Confucius engages in relatively conventional rational or formal reasoning, and issues relatively direct instructions. Laozi, the attributed author of the Dao Dejing, was engaged in something more like meta-rational or post-formal reasoning.

As I mentioned before, language divides the world up into categories. How can you make a small text which still transmits meaning?

One method is to narrow the scope of your discussion. Write 1,000 words about dogs, and you can say a good deal about dogs. Write 1,000 words about the whole world, and you won't be able to say much about dogs specifically.

Another method is to allow multiple interpretations, and let the reader do the work of sorting out what you mean.

Yet another method is to use specialist vocabulary, which allows you to combine repeated information into a shorter representation (or repeated ideas into words instead of whole passages).

You can also try assuming a greater background knowledge from the reader – for example, 1,000 words written about dogs for veterinarians could get into quite a few details that a general introduction to dogs would not have space for.

The point I'm making with the above considerations is that there is a trade-off between the amount of work the writer has to do, the amount of work the reader has to do, the length of the document, and the amount of meaning transmitted.

The Dao Dejing concerns general patterns in the world. It works at an extremely high level of abstraction. In about 7,800 words split across 81 chapters (or roughly 96 words per chapter), it attempts to describe the entirety of the general nature of the world. It is highly compressed. It relies on knowledge, life experience, and consideration to process.

For many years, I avoided reading the Dao Dejing. It sounded too mystical. How could a method or practice in such a text work reliably?

I didn't realize that the Dao Dejing was different from conventional mystical works, even though people have sometimes interpreted it in a mystical way. It is less mystical than it is mathematical. It is based on human nature, epistemic virtue, and dimensionality.

3.a. Notes on Dimensionality

The most important thing to realize is that the world is thick, heavy, and substantive, while our models of the world are light, thin, and insubstantial.

A boulder might weigh two thousand pounds, while the hardware of a computer simulating that boulder in a video game might weigh two pounds, and the simulation itself may be lighter still, consisting of tiny electrons in a low-friction environment. The boulder is heavy and round, and takes a great deal of energy to move. Our simulation is small, flat, and light, like a piece of paper, and thus takes very little energy to move. The boulder is composed of, perhaps, trillions of molecules, while our video game simulation is likely just a few hundred polygons stitched together.

The boulder is the RMS Titanic. Our simulation is a papercraft model of the RMS Titanic.

We could sink a papercraft model of the RMS Titanic 1,000 times without losing a single crewman. It's just a (figurative) piece of paper.

A boulder in real life is a volume composed of molecules. A boulder in a video game is an empty solid bounded by polygons. In real life, the material of a boulder has many options for composition at any given location. In a video game, it is reduced to two: { empty, solid }. This narrows the dimension of the boulder material.

We can conceive of any characteristic of the boulder as a dimension. It isn't just a matter of 3 dimensions of space. The ability of the boulder to break, and the number of ways it might break, could be a dimension. Video game boulders generally cannot break, unless they have been specifically designed to break, and even if they can, it is generally only in a single, predefined way.

All human models must reduce the number of characteristics (or dimensions) considered, and reduce the number of options per dimension (scope of a dimension) considered.

Or to put another way, any model that a human being devises is necessarily less complex than the world.

Suppose that we attempt to map the footprint of all the buildings in an entire city using a grid of 5 meter units. (Think Minecraft, but with bigger blocks.) This might be a pretty good approximation. However, there will be a physical distance between the edges of a block on our map, and any actual building. If the building happens to line up pretty well with our grid, that could be small, just a few centimeters. If it doesn't, it could be the size of an entire sidewalk.

Does that matter? It depends on what we do with it.

If we make a nice decorative overhead map to place in the city hall, it's possible that no one will notice. If we give this map as guidance to an autonomous vehicle, it might accidentally run people over.

3.b. A Note on Perspective

Human beings generally see the surface of things, not their internal volume. Thus, even in creating a model, it may be based on the appearance of an object or phenomenon, and only a best guess at its substance.

3.c. The Dao

How is it that, in an attempt to take control over something, one might lose their grasp over it? Among other reasons, in applying more force to the low-dimensionality model in the high-dimensionality space, the amount that the distance between the model and reality matters will increase. It may be possible to exert a small force on it with a small amount of knowledge, but exerting a large force may cause unexpected behavior.

How do you deal with this? Let other people do things. Let other people, who are closer to the situation, make decisions. Let things go if you can't control them, rather than hurt yourself or others by trying and failing. Accept uncertainty. "Act through inaction."

How is it that something which is an apparent contradiction is not a contradiction? Viewed in 2 dimensions, a coin might appear to be only heads, or only tails, but viewed in 3 dimensions it is a single object with two main sides, and there is no contradiction.

Laozi wants the reader to step back and view things from a higher level of abstraction, to step back and not exceed their planning and modelling horizon, to step back and allow others to do the labor of living and deciding.

The teaching of Laozi is less like a story of the supernatural, and more like (although not entirely like) Big-O notation. Computer scientists like to track how many units of resources a computer algorithm will use not through a highly-detailed representation, but by its fastest-expanding term as a function of input. If loading your favorite video game will take (n^2 + n^3) minutes per gigabyte for some reason, Big-O notation focuses on the n^3 and ignores the n^2, so it is O(n^3). This will be wrong for small numbers, and approximately correct for large ones.

The truth of the Dao is statistical. The greater the mismatch in dimensionality, the greater the wisdom of not-doing.

This is not the whole of the Dao, and it is far from the only interpretation.

4. A Politics of 4-Dimensional Knowledge

At the core of a (political) belief, there is a { concept }. This could be something simple. For example, "Adult male grizzly bears typically weigh about 800 pounds."

We could treat this concept as a binary. Either the bears weigh about 800 pounds, or they don't.

However, what if we want to consider a second idea, such as, "Adult male grizzly bears typically weigh about 300 pounds"? Now we have a contradiction! The bears can't "typically" weigh both 300 pounds and 800 pounds (even if individual bears may vary).

Rationalists like to make annual predictions, and in an annual prediction you record both a concept and the percent chance you think it could happen. This is two-dimensional: { concept, probability }.

Let's suppose that we live near Yellowstone National Park (in Wyoming). The park office calls us and asks us to help transport a tranquilized grizzly bear. Due to a bureaucratic screw-up, the park office doesn't know the size of the bear. How big of a truck should we bring?

Let's say that we text message two hunters whom we trust about how much grizzly bears weigh. Dale, an experienced hunter who lives on west coast, says they weigh about 800 pounds. Gred, an amateur hunter who lives in North Dakota, estimates that they weigh about 300 pounds. We also hear from someone's conspiracy theorist uncle next door, who claims that "5,000 pound Africanized grizzly bears" are terrorizing the country.

{ 5,000 pound bear, 0.01% } { 800 pound bear, ~70% } { 300 pound bear, ~30% }

This approach allows us to hold two or more ideas simultaneously. This allows us to make policy that covers multiple cases, and adjust our investment or risk for the probability of each.

Our estimated odds of an elephant-sized grizzly bear are low, and a truck to carry one would be specialized equipment. Our best bet would be to bring a truck capable of carrying an 800-pound bear, which also nicely covers the 300-pound bear case.

However, while the probabilities of our estimated bear weights may sum to 1, this isn't really sufficient to describe the situation. Neither of the hunters actually live near Yellowstone. Maybe Yellowstone grizzly bears are a different size.

{ unknown pound bear, 50%} { 5,000 pound bear, 0.01% } { 800 pound bear, ~35% } { 300 pound bear, ~15% }

One approach would be to just throw in another concept to represent our lack of certainty.

A prediction market has a way to measure certainty: how much you are willing to invest. As such, we could try quantifying certainty with a third term: { concept, probability, certainty }.

{ 800 pound bear, 70% chance, low certainty }

We could try representing certainty in a more quantitative way, such as with a percentage. We can also think of certainty in a more qualitative way: "How much effort did I put into finding out if this is true, what sort of information did I have access to, and how good am I at this sort of investigation?"

Now, let's add one more term: { concept, probability, certainty, investment }.

Just how much weight are we putting on this idea? If the idea is, "I tied this rope harness for rappelling down this cliff correctly," then you're betting your life on it. If it's, "This TV channel will have something good on it," then you're probably not very invested in it at all. Investment in this sense can be moral or social; it doesn't have to be financial.

In real life, no one carefully tracks their political beliefs in a spreadsheet with four terms. The reason I bring up this model is so that we can see the flow.

If we want to increase our level of investment, then we should increase our certainty.

If we have a high level of certainty, then we can increase our investment later.

If concepts are too distant or incompatible, and there is a broad spread of probability among them such that no concept is dominant, then it may be difficult to make policy.

Obviously, I also don't keep a spreadsheet of all my beliefs with 4 terms each. This is just how I generally approach knowledge. In less certain areas, I carefully feather posts with qualifiers to mark uncertainty, move back a step in abstraction, or speak of real things as though they were hypotheticals.

I've heard that I have been described as so unknowable that I am, "Like a fae creature." I've found that it's often difficult for large language models to classify my writing as either "left-wing" or "right-wing."

I think what may be tripping some people up is that, while I am a compassionate person, I treat political beliefs as a portfolio.

I design or endorse policy based on what I think the relative probability of different political facts is, generally thinking in terms of policy that is reasonably effective in more than one case. This may mean moving up a level and making policy that is more abstract, or which works on higher-order effects, or which includes conditionals.

5. Addressing Earnest-Peer's Criticism

Earnest-Peer wrote:

Miti at some point got it into their head that they should roll out their whole worldview at every opportunity; [...]

This is the simple one to answer.

The longer that I write, the more ideological basis I develop that is either unique to Mitigated Chaos, or which is farther away from local background knowledge. Earnest-Peer is correct that the blog now repeats concepts more often.

I discovered that a number of people were having more difficulty understanding my posts than I expected, so I decided to provide more context to make my posts easier to understand.

I discovered that Tumblr's current readership are younger and less experienced than I expected, so I decided to provide them with more relevant background information.

A broadly-circulated Tumblr post will go outside of its local social context, and thus has to be understandable to people who haven't read the long history of posts on Mitigated Chaos. (I have roughly 8,500 posts.)

Posts to Mitigated Chaos are not only written for the benefit of a small circle of discourse hobbyists, even though those are the main people who respond to them. They are meant to provide ideas or information, as well as example behavior, to a larger audience of passive readers. The re-emergence of discursive norms (such as the idea, "ad hominem is a fallacy,") on Tumblr is likely due, in part, to the long-term behavior of discourse blogs, many of whom faithfully upheld discursive norms even when this was socially disfavored.

Some of this is also similar to writing formulas at the top of a math sheet so as to make it easier to remember the terms. (This most evident in the response to Oligo*'s post "economic disextensification.")

6. Addressing GhostPalmTechnique's Criticism

On November 3rd (just days before Donald Trump won the 2024 Presidential election), GhostPalmTechnique wrote:

Did mitigatedchaos used to be more intelligent and less racist, or did my patience with them just wear out completely?

Based on the essay contents above, I think we can safely dismiss the idea that I've become much less intelligent. The blog does repeat itself more often, but this is in order to be better understood, as outlined above.

6.a. On Methods

There are two ideas that I am particularly conscious of in my writing. The first is compression, by which I mean here the way that humans tend to remove or lose information that's part of an idea when they remember it. The second is source amnesia, the tendency of humans to forget where they heard some fact or where it came from.

The first tendency may result in information getting mangled. The second tendency may result in misremembering fictional information as fact.

I will write something like, "Suppose that Norwegians are, on average, one inch taller than Italians."

Suppose that someone misremembers this hypothetical as a fact.

Height is generally considered a favorable, but relatively morally neutral trait. (It is considered less morally weighty than, say, obesity.)

A one inch difference in height is not particularly noticeable at the population level. There could easily be a 6 foot tall Italian and you wouldn't even think about it.

While there has been a long history of violence on the basis of ethnicity, someone would have to be a real fucking psychopath to massacre a population based on a one inch average height difference.

Within the United States, "Norwegian" and "Italian" aren't currently categories subject to major political dispute. (The broader category of "European" may be subject to political dispute, but this is just two different groups of Europeans.)

Suppose that someone misremembers the content of the hypothetical. He asks someone, "Did you know that Frenchmen are, on average, one inch taller than Italians?" Again, similar considerations apply.

Additionally, this is a hypothetical. You haven't actually learned any information about the relative heights of Norwegian and Italian people, including whether they even noticeably differ, at all. Italians could actually be a full six inches taller, and you'd have no idea.

We will get to why I decided to use such a hypothetical at all, later. However, every element of this hypothetical has been very carefully crafted to prevent it from being useful in, or promoting, ethnic conflict.

So, as a first tactic, we can make a hypothetical harmless if someone forgets that it's a hypothetical.

As a second tactic, such as in the post about cryogenic lunar prison which was designed to assess Tumblr users' intuitions about imprisonment, we can insert an element into the hypothetical that is obviously not fucking real, so if someone remembers the hypothetical but forgets that it's a hypothetical, the presence of a unicorn or whatever in the hypothetical will remind them.

6.b. On Mitigated Chaos

This work is the result of careful and deliberate effort, but it is invisible to those who deem Mitigated Chaos to be "reactionary."

Mitigated Chaos is closer towards the natural center of government, and closer to the edge of contemporary ideological development, than such people give this blog credit for.

The theory that the progressive wing of the Democratic party has reduced the agency of Democrats, that they are too obsessed with consensus, resulting in a stifling intellectual atmosphere that reduces the ability of the party to adapt and change?

Yes, that's the theory of Mitigated Chaos. It's also the theory of reasonable centrist Nate Silver, and hardcore partisan Democrat Matthew Yglesias. And I pointed out this problem on election night.

Right and left are methods, tactics, schools of thought, worldviews. Each collection taken as a whole, they are rich, but determining the best frame or technique to apply to a given situation requires the labor of judgment, discernment, and observation. (It also requires the humility to recognize when one is wrong and the courage to act when one is right. These values are necessarily in tension.)

As I wrote to sophia-epistemia:

Rather, much as if virtue were always rewarded, we would not need to praise it, it is not the law itself which is the nature of good. ... Due to the low-dimensionality nature of law, one cannot simply outsource one’s judgment to the law. One must actually be good.

Ideology is not a substitute for being a good person.

6.c. On Justice

Likewise, if virtue were always rewarded, we would not call it virtue. There would only be productivity.

What differentiates oppression from mere human suffering is that there is an oppressor.

What is it that differentiates justice from mercy or compassion?

Justice is taken in response to agentic action. It concerns an action which was taken, or a rightful duty which was not upheld. Let us use an example to consider the nature of justice.

Suppose that there is a dam overlooking a farmer's field. The role of maintaining the dam belongs to a local official, who is compensated for his efforts. One day, shortly before the scheduled harvest, a storm comes and the dam breaks, washing away the farmer's crop.

If the local official upheld his maintenance duties and the dam broke anyway, then this is ordinary human suffering. The local official may compensate the farmer, but this is morally praiseworthy, not morally obligatory.

If the local official neglected his duties, then the official compensating the farmer for the lost crop is justice.

If the local official was unable to fulfill his duties, and refused to resign, then the official compensating the farmer is justice.

If the local official was unable to fulfill his duties, and attempted to resign, but the central government refused to allow him to resign, then the local official compensating the farmer is not justice, and the central government is the guilty party that owes compensation.

If the local official was unable to fulfill his duties, and attempted to resign, but there was no qualified person to replace him, then the field being destroyed is once again mere human suffering. The local official should perhaps pay some amount of dues in compensation, but the fund is now merely an insurance scheme.

One purpose of justice is to change the nature of the world. If we commit to the rule that murderers are to be imprisoned, and then we reliably follow through by catching and imprisoning murderers, then murderers exist in a world in which murderers are reliably imprisoned.

A significant chunk of opposition to imprisoning murderers can be thought of as moral alternation – switching between more coherent moral systems so as to reach an outcome which is not approved of by any of them.

Under a utilitarian moral view, it may be acceptable to imprison someone who is at an extremely high risk of committing a crime, even if he has not yet committed that crime, and to release someone who is at an extremely low risk of committing a future crime may be appropriate even if he did commit a crime.

Under a more rules-based moral view, it may be unacceptable to imprison someone who is at very high risk of committing a crime if they have not committed the crime, while if they have committed it then they are deserving of punishment.

Both of these views are coherent and have means of addressing crime.

Under moral alternation, someone may switch halfway through. First, they would say that the high-risk potential murderer should not be imprisoned because they have not committed an actual crime yet (rules-based morality). Second, after the crime, they would state that punishing the criminal will not bring the victim back to life, and will just create more suffering (utilitarian morality). This negates the value of the victim's life in both cases, effectively saying that the victim's life doesn't matter.

The concept, "not guilty by reason of insanity," concerns individuals who are insufficiently agentic, who don't just set aside incentives, but who are actively incapable of responding to them.

If someone engages in an action that they know will cause them to become a violently insane, and then engages in crime, then if they should ever stop being violently insane, then they are guilty and should be imprisoned. (If they remain violently insane, they should be separated out from potential victims.)

If someone becomes violently insane by accident, and then engages in crime, then this is mere human suffering, but they should be kept physically separate from other potential victims until they are no longer violently insane.

If someone becomes violently insane by accident, and a local district attorney deliberately or negligently releases him into the general population, and the insane man then commits crime, then the insane man is innocent but should be kept separated, while justice requires that the local district attorney pays compensation to the victims.

Human beings do not all agree on what is fair. The world itself, however, is not fair. Someone can 'do everything right,' work as hard as they can, help others, and still be destroyed by random chance. Some people may commit no crime as generally understood, but still die from a genetic mutation at a young age.

We use justice to make the world more fair, but not everything we do to make the world more fair is justice. We have different words for mercy and compassion because they have a different causal nature.

6.d. Regarding Tension

Social Justice seeks equality of outcomes between demographic groups. I don't believe that supporters of Social Justice will dispute that description.

There are many people out there who have trouble imagining things beyond their own life experience. People flatten things. They forget information. They overgeneralize from a small number of examples. (Some people say that they are progressive and then flatten everything out anyway.) Liberal left-identitarianism can act as a counter-force to these tendencies, when it is rooted in truth.

Human beings are both similar and different. They have differing life experiences. They read different books, and watch different movies. They have differing personalities, and may come to different conclusions from the same information. People understand others in part through understanding themselves – yet they are not like others, and it takes a great deal of time and effort to cultivate self-knowledge, when it can be done at all. Thus, human beings fundamentally do not understand each other.

This is a fundamental tension within human existence.

Tension is not inherently bad. Tension is necessary for human existence.

We are a "resolution" or "answer" to multiple evolutionary forces acting in different directions. Multiple simpler forces acting in opposition can produce a more complex, higher-dimensionality result than any of those forces acting alone. We are high-dimensionality beings because we must solve high-dimensionality problems.

Remove too much tension, and human dimensionality would collapse.

People are always trying to solve humanity, to devise a solution which will resolve all conflict and make humanity "happy" forever.

Utopia is a featureless flat plane. No human beings live there.

6.e. Regarding Monster Trucks

Social Justice is distinguished from conventional liberalism through seeking "equity," or equal outcomes, rather than "equality," or equal procedures.

During the late 20th century and early 21st century, many people thought that human beings were sufficiently similar to each other that equality before the law, in combination with things like universal education and conventional human rights, would gradually equalize outcomes between demographic groups.

The coalition of supporters for equality before the law in that period included both those who thought non-discrimination would close group outcome gaps, and those who thought that legal equality would not close group outcome gaps, but that it was good anyway.

For example, suppose that a monster truck rally only accepts male monster truck drivers. A right-liberal might believe that interest in monster trucks is split 90%-10% between men and women, so that the "natural" rate of female monster truck drivers is only 10%, yet support formal equality in allowing women to participate as monster truck drivers anyway.

I want to stress: this is not a contradiction. The right-liberal sees monster truck drivers primarily as individuals, and not as representatives of identity groups. The right-liberal can say, "Women vary. Women are diverse. Some women are different from other women, and may enjoy driving monster trucks even if other women don't."

There is a critique further to the right that allowing female drivers will start a gradual shift where people treat monster truck drivers differently, leading to a gradual feminization of the profession and loss of male interest. Nonetheless, the right-liberal position above is fairly well-hedged. Not only does our right-liberal allow for individual variation, but he also allows for the fact that he may not know the "correct" gender ratio. I have personally known one woman with a disability who enjoyed watching professional wrestling – the right-liberal above does not need to dispute the legitimacy of her interest, because he is not taking a position on her interest.

A left-progressive, a supporter of Social Justice, would see anything other than a 50%-50% male-female split (or otherwise proportional) as evidence that either something was wrong with the monster truck rally, or that something was wrong upstream of the monster truck rally.

Contemporary left-progressives generally believe that those who are in "privileged" or "dominant" groups have agency, while those who are in "marginalized" or "oppressed" groups do not. As such, if a monster truck rally audience is 80% male and 20% female, while visitors to a quilt show at the same time are 80% female and 20% male, a contemporary left-progressive will not say, "If the women are at the quilt show, they cannot also be at the monster truck rally." They also won't say, "This quilt show is problematic, and needs to figure out how to attract a larger male audience."

The left-progressive will say that it is the monster truck rally that is problematic and needs to change.

The right-liberal does not place himself in conflict with women who love driving monster trucks. The left-progressive places himself in conflict with men who like watching monster trucks.

In the Dao Dejing, Laozi counsels to acknowledge the sufficiency of sufficiency. Between our hypothetical right-liberal and hypothetical left-progressive, who is in greater accordance with the Dao?

6.f. On Social Justice

How is it that Social Justice is justice?

Justice concerns the response to both agentic actions and failures to uphold duties.

Social Justice defines equal outcomes as the default state of the world, and defines deviations from equal outcomes as either the result of agentic action or a failure to uphold duties.

How is it that Social Justice is social?

Social Justice concerns a perceived net balance of harms between demographic or identity groups. (It is social justice and not individual justice.)

Let's lay out a problem and discuss the limits of the social justice and individual justice frameworks.

Suppose that French Canadians are considered to be a marginalized group. (They generally aren't considered such in the United States, but imagine that they were. Maybe Canada has a sovereign debt crisis and breaks up in a violent civil war or something.)

A French Canadian high school student goes to a school in New York. There, 30 American students mock her accent when speaking English and insult her. The rest of the student body (470 students) remain quiet. This leaves her feeling unwelcome and alienated.

From the perspective of individual justice, there are 30 individual cases of mild social attack against the French Canadian student, and the proportional response would be a mild social attack towards each of the 30 insulting students individually.

One motivation for the adoption of "Social Justice" is the aggregation of smaller effects or harms. It is difficult to gain social leverage on 30 other people at once. A one-on-one social dispute pits one person and their social abilities against another person and their social abilities. A 30-on-one social dispute pits one person against either 30 single encounters in a row (if the opponents are not cooperating), or against a formation that can leverage the talents of different members (if the opponents are cooperating).

From the perspective of social justice, all 30 insults are aggregated into a single charge ("anglophone supremacy"), and the quiet students have failed to uphold their duty to restrain or criticize the insulting students, so a single case is opened up against the whole student body (all 500 students are "anglophone supremacist").

This allows a different path. If the insulting students are socially adept, while many of the quiet students are socially weaker, it may be easier to shift the burden of moral liability from an insulting student over to a quiet student in order to extract compensation. Alternatively, because the entire student body are considered morally liable, a punishment or demand for compensation may be made against the entire student body, without having to do the work of proving the involvement of each individual.

It's easy to see why this is tempting, but it has several problems.

For one thing, although the student body may have an informal social hierarchy, it doesn't have a formal one like a conventional military in which there are officers with enumerated powers to enforce compliance. (No high school clique on Earth has the legal power to, for example, shoot deserters.) Most of the quiet students may have no leverage with which to influence the insulting students, and may have never suggested insulting the French Canadian student in the first place. Spreading out the punishment wastefully hurts innocent people, but assuming that the balance of harms is maintained, it must also be much weaker (spread over 500 students) than a more focused punishment against individual wrong-doers (spread over 30 students).

There appears to be a widespread misconception among left-progressive supporters of Social Justice in which they assume groups like "men" are much more organized and coherent than they actually are.

Second, viewing the interactions through the lens of a collective balance of harms encourages a mentality of conflict and grudge accumulation. People will be more likely to interpret ambiguous situations in a negative way, rather than attempting to either defuse them or reduce the ambiguity before making a judgment. Since people can make mistakes and flubs, a little positive interpretation can help smooth over social interactions and prevent the emergence of conflict where it would not naturally exist. Additionally, in a balance of harms framework, someone has to accumulate and maintain a collection of grievances to bring out (on demand) to prove that they have been harmed more, in order not to be punished. They can't forget about them. They can't set them aside. That means losing leverage to not be hurt!

Some left-progressives would attempt to wring all the nuance out of the previous paragraph. "Oh," they would say, "You're saying that people should accept being harmed and just forgive everything and not respond?" Such people should learn more humility. They tend to assume that they know a lot more about what causes which outcome, and to what degree, than they actually do. They want what is easy, not what is right.

Third, assume you are one of the quiet American students. The insulting students keep racking up harms, and you keep getting punished for it. Which is easier: getting the insulting students to behave properly, or driving the French Canadian student from the school by joining the pile-on?

Some readers would say that the latter option is immoral, and they would be correct. The problem with collective punishment in this sense is that removing the punishment is a gain, so the collective punishment is effectively rewarding people for acting immorally.

People are capable of heroic acts of moral sacrifice, but they will generally only make them under heroic circumstances, and not all of them will. It is best to limit the amount of moral sacrifice that people must make, in order to save it for when and where it is most needed. (Confucius would agree, here: "[The higher type of man] might be imposed upon, but not utterly fooled.") One way to improve the world is simply to make it easier to do what is right.

It is good to be kind to the French Canadian student. The uneven suffering of the French Canadian student would matter.

However, we need a different name for "being kind to the French Canadian student." Justice, especially collective justice, is not the appropriate frame to apply to the problem.

Justice is about healing, yes. In our example with a local official and a broken dam, the payment from the local official helps to make the farmer whole for the value of the lost crop. However, justice is also about imposing a right ordering on the world, and for that to work, it must be applied carefully. Punishing random people does not create the right incentives, and causes individual injustice.

Justice is work. To effectively carry out justice requires personal development. You must know yourself in order to know truth, and know yourself and others in order to exercise sound judgment.

6.g. The Tortoise

One way that people try to control others, consciously or otherwise, is through attempts to reduce their dimensionality.

By limiting the range of people's responses, they can be predicted. By being predicted, they can be controlled. By being controlled, their actions can be directed towards some ends. By directing their actions towards some ends, they can then be exploited.

If people didn't limit their behavior at all, and might act like a saint, or might act like a serial killer, at any time, randomly, it would not be possible to have a society. Set aside modern society – if ancient agriculturalists did not control their own behavior and therefore plant, harvest, and store food, they would die.

It is wise to work as a subordinate for a larger force, and also it is wise to maintain your autonomy. Sometimes, it is good for humanity to cooperate. Other times, it is good for humanity to compete. (One reason that markets are such a powerful force is that they combine both of these elements.)

Instincts are robust, but inflexible. An animal doesn't need to be taught that which is instinctual, so it might hatch from an egg alone, survive, and still prosper. However, if an instinct is not in accordance with the environment, then because it is biologically rooted, it is very expensive to change.

Learning is fragile, but adaptable. Some animals learn by watching other animals of the same type. If no other such animal is around, then the information will be lost. However, if a different behavior comes along, then the new behavior is easy to learn.

This creates a spectrum.

An organism is served by having a range of flexibility to change.

The more difficult some element of an organism is to change, in general, the more this element should be general and timeless, applying to a very broad range of conditions. The easier some element of an organism is to change, in general, the more specific and timely it should be, applying to conditions in the moment.

Thus, the tortoise has a hard shell, flexible legs, and in his little head, a tiny brain.

There are some people who wish for humanity to be a blank slate, with no natural differences in personality.

They believe that this will allow them to 'solve' humanity, and create a utopia where there is no suffering or conflict. All that is to be done is to devise the right ideology and then scour away all the self-perpetuating cycles of oppression inherited from human history.

Suppose that humanity really were uniform in temperament and, with the exception of age and injury, in capability. There is no one so stubborn that he might refuse an ideology out of spite, and no one so stupid that he might not understand it. Ideology rolls smoothly across humanity like a wave, and no one contests it.

What happens if it's a bad ideology?

In this case, it might propagate to the limit of the communications network, and wipe out an entire population. If recorded in text, it might propagate forward in time, and wipe out successive populations until further understanding is lost.

So, is it good if humanity is a blank slate?

No.

A world in which no one has natural resistance is a world in which complete victory is possible, a world in which there is no one left to fight. It's also a world in which complete loss is possible, and its advocates know this.

A world in which there is natural resistance is a world in which complete victory is impossible, and there is only a favorable balance of forces which must be maintained through conscious effort.

A humanity in which there is a broad range of natural variation in personality, in which some people are grouchy while others are optimistic, in which some people are violent while others are saints, is a humanity that is more resilient.

Which is good, structure or flexibility?

It is good to have structure, and it is good to have flexibility.

Which is good, order or chaos?

The limit of order is a one-dimensional point, and the limit of chaos is noise. It is good to have order, and also, it is good to have chaos.

7. Is Mitigated Chaos Racist?

The above section is one of the most powerful general arguments for the diversity of humanity most people will have read. It cuts across almost every dimension of human existence. Arguments more powerful than this will generally be drawn from individual readers' personal life experiences.

It is also the argument of someone who understands that diversity exists at multiple levels of abstraction. What is diverse at one level may be homogeneous at another, and what is homogeneous on one level may be diverse on another.

7.a. Mistakes and Conflict, Fast and Slow

Both left and right apply mistake theory (the theory that conflicts are mostly due to misunderstandings and mistakes) and conflict theory (the theory that conflicts are mostly due to differences in interests) to the theory of evolution. Mistake theorists are more likely to assume that an opponent is arguing in good faith. Conflict theorists are more likely to assume that what an opponent is saying is enemy propaganda, or at best subconsciously selfishly motivated.

It is not correct to be entirely mistake theorist, because sometimes people really do maliciously lie. It is also not correct to be entirely conflict theorist, as this can trap someone in a cynical information bubble and create conflicts that did not need to exist. (To be fully either one or the other is mentally disordered.) It is the approach of mistake theory that creates lift. A careful application of conflict theory can reduce drag.

Broadly, when it comes to politics, there are two theories of evolution. Either evolution has an effect on humans in the short term (fast evolution), or it only has an effect on humans in the long term (slow evolution). (There are also religious individuals who do not believe in evolution.)

In general, fast evolution is considered the right-wing position. It implies that humans are more varied. If humans are more varied, then in the absence of environmental pressures forcing convergence, their lineages will be more varied, too.

In general, the left wing believe in slow evolution, which implies that humans are more similar to each other. This may even extend out to animals, especially mammals. Some left-wing individuals may believe that animals are very similar to us, or that all beings with experience are very similar to us.

These stances can be viewed as expressions of the general right-wing mindset of an expectation of scarcity, and the general left-wing mindset of an expectation of abundance, respectively. (In real life, evolution is more complicated than this.)

Often, the left believe that support for the fast theory of evolution is driven by a desire to feel superior to, dominate, and exploit others. Some believe that this is a personal character flaw, while others view it as a self-sustaining phenomenon of self-replicating ideas. This latter view is one reason so many leftists are so eager to censor what they describe as "hate speech."

Often, the right believe that support for the slow theory of evolution is driven by envy, resentment, and greed. The theory is that people shift what they perceive as their own flaws or weaknesses onto others. For example, someone who is lazy might prefer to believe that a diligent worker gained their position unfairly rather than acknowledge their own laziness. Alternatively, perhaps even more cynically, supporting the theory of slow evolution where humans are all equal may be a means for lazy people to band together to selfishly transfer resources away from hard workers.

It might be shocking for left-leaning people to see the right-wing position presented as such a direct inversion of their own ideas like that.

The reality is that both views may be the result of selfishness, and both views may be the result of not wanting to be blamed and "held accountable" for the actions of others. The motivations, or perceived motivations, are not sufficient to establish the truth.

7.b. The Conventional Wisdom

In my opinion, left-leaning people are very used to seeing conventional Western right-racism as a selfish rationalization. After all, from the inside, an emotionally-motivated selfish rationalization might feel like truth.

How do you differentiate an emotionally-motivated selfish rationalization that you hold from truth? Well, for one thing, you argue with people in order to encounter arguments that you haven't heard before and that you might not think of yourself due to emotionally disfavoring them.

Another way to deal with it is to create a predictive model, and then go out and measure something to see if the prediction fits.

The problem for race in this respect is that, due to the expectation of conflict, it is a political-epistemic event horizon, similar to the information environment around a major war where multiple opposed actors plus a bunch of other people making stuff up are all acting in opposition to each other, making it very difficult to ascertain the truth. Putting in enough effort to get to the bottom of the mystery of race suggests that someone cares way the fuck more about it than any normal person should, which makes their judgment suspect to everyone else – by conducting the research, they can no longer report back their findings to people who haven't done the research.

This is different from, say, guided busways. I can write to you about guided busways (such as this one in England, isn't it neat?) because as far as anyone is aware, very few people have been killed over guided busways.

So I didn't get to the bottom of the mystery of race. I just made bets based on observing the debris field around it. Once you understand this, my positioning becomes much more obvious.

Support for race-neutral standardized testing and credentials disaggregates individual performance from race, making it less rational for e.g. patients to care about the race of their doctor as a proxy for their doctor's talent, skill, or medical school admission criteria.

Support for arresting and imprisoning violent criminals regardless of race reduces the likelihood of being victimized in a cross-racial violent crime, and we should be arresting violent criminals anyway.

A positive, but modest amount of highly-filtered immigration keeps things fresh while not wastefully causing resource conflicts between natives and immigrants.

A cross-racial civic nationalism that emphasizes the strengths and virtues of the country while also selecting figures from minority groups to celebrate helps people feel included and not resentful.

Race-neutral anti-poverty programs such as free public education and healthcare for the poor are more like social insurance than extortion, even if they disproportionately go to members of racial groups who are currently disproportionately poor, because a) any citizen could conceivably access them if they fall on hard times, and b) if the balance between groups shifts later, the programs themselves automatically shift.

Social norms opposed to majority identitarianism and in favor of very mild liberal left-identitarianism should prevent too much cultural flattening.

This is a staggeringly well-hedged portfolio. If the left-wingers are correct, then outcomes should converge across groups over time without having to have a massive confrontation. If the right-wingers are correct, then damaging side effects are eliminated or controlled, which should prevent the situation from boiling over into a massive confrontation. Along the way, we reward hard-working and talented individuals, and don't needlessly make people people feel bad about their race.

What about the project of lifting up the whole world?

We already had that. It was called the United States of America, a continent-spanning liberal-democratic superstate with a constitutional right to freedom of speech, which secured the world's sea lanes for free trade with the world's most powerful navy.

With all bottleneck resources being sold on the global market, advanced or industrialized countries didn't need to put together armies in order to establish colonies in order to get bottleneck resources to improve their armies so that they could resist the armies of the other industrialized countries doing the same thing.

OK, well what about the Iraq War? Doesn't that undermine global peace? Surprise! I don't like George W. Bush.

7.c. Impatience

Why patience?

The more I try to force the issue of development, the more I have to control people. The more I have to control people, the more morally liable I am for their actions and what happens to them, and the more careful I have to be about being correct in order to avoid hurting them.

It's better to create a favorable environment for people to improve themselves, reward the virtuous, and punish the worst criminal elements, rather than try to strictly control everything.

Consider the program I have laid out. In theory, it should only fail under relatively extreme conditions – conditions that would tend to make themselves obvious.

Through this method, I could avoid taking a hard position on race, which meant that I could avoid researching race, which meant that I could mostly avoid thinking about race.

A contemporary left-progressive would take a conflict theorist stance on this. They would argue that (although I have not disclosed my race), I was merely wallowing comfortably in racial privilege. That the machine suppresses majoritarian racial organizing would not register for them; they would consider that merely a selfish rationalization.

In 2014, one of the major left-leaning outlets in the United States, NPR, started having opinions on the 'correct' racial makeup of engineering departments. But that was just one news outlet. It might just be clickbait.

In 2019, five years after left-progressives started loudly demanding that I take race seriously, and that I think about race, and telling me that race was the single most important issue facing America, I decided to take race more seriously. I would no longer delegate thinking on the matter based on the expectation that the Democratic party would follow an expert academic consensus, as I had earlier.

Rather than researching the deep mystery of race directly, I researched it at one remove – I went looking for social programs that would improve racial outcomes. If I found good ones, then I'd be able to redirect the radicals into doing something worthwhile. I didn't find much, and what I did find wasn't especially effective.

The behavior of left-progressives in America in 2020 looks reasonable if you assume there's a long list of highly effective social programs that just aren't being implemented due to political opposition. However, if you actually go looking for this list and don't find it, then their behavior looks, let's say, a bit less hinged. Left racial activists would memorize impressively long lists of racial grievances which normal people, who weren't expecting to have an argument about race, were not in a position to counter.

(It wasn't all bad. It was a tremendous opportunity to learn more political theory through observation, and in much safer circumstances than those suffered through by the thinkers of the 20th century.)

However, left-progressive activists aren't the entire Democratic party. There was a question, then – did they actually have power, or were they just really loud?

During the pandemic, there were institutional attempts at racial healthcare rationing [The Atlantic]. Well, okay, technically, a hospital system is not the federal government, and it's not the Democratic party. Maybe a few administrators just went a little crazy?

After the Biden administration came in, they attempted to do 'race conscious' farm aid [Reuters], and had to be shut down by the courts.

Now, a lot of people who are affiliated with the Democrats would attempt to downplay this. They'd say it didn't go through, so it doesn't matter. Of course, it shows that the intent is there, the intent is strong, and the intent isn't going away on its own. They'd say that it got shut down by judges. Well, who appoints Republican judges? Republican politicians! By comparison, what do law students in left-leaning colleges support?

Reuters and The Atlantic are not "right-wing" sources. I could build up a larger database of this kind of misbehavior if I wanted to. It wouldn't matter.

The Democrats have broken each pillar of the program that I laid out, and they did so without the kind of hard evidence that would justify doing so.

Kamala Harris might not have campaigned on it in 2024, but she didn't actually campaign on much of anything at all, and she certainly didn't campaign against it.

So yes, I'm more aligned with the right-wing currently.

Am I racist?

7.d. Mitigated Present, Mitigated Future

Long-time and close readers of this blog may have heard of Kymriah, the first FDA-approved commercial gene therapy. It was approved in 2017. More gene therapies have followed.

According to drugs.com, Kymriah costs $612,000.

Back around the year 2000, the United States paid about $2.7 billion for the Human Genome Project to sequence one human genome. Now, genome sequencing costs less than $1,000.

If someone wants to call my carefully hedged set of bets "racist," then they can, I guess. Anyone can cook up their own bespoke definition of "racism," or borrow one from whatever activist they like. Anyone can choose to ignore the context and care I put into my work.

It'll mean they're still mentally living in 2016, though.

It's possible that the genetics industry is going to stall out, that the relative dimensionality and coherence of the computer-backed human medical industrial system will crash against the dimensionality of the millions of years of evolution of the human body, spread across trillions of microscopic cells, and fail to gain traction. It's possible that they'll only cure a few rare diseases and the price will stay stuck at $1 million dollars indefinitely.

It's possible – but is it a good bet?

I wouldn't delay having children in the hopes that the genetics industry would give you the perfect baby – there are ethical questions with that, anyway. I wouldn't abolish public schools on the theory that in five years, every child's gonna be a Mozart or a Newton or an Einstein – that would also raise even more dire ethical questions (as implied by the rest of the essay).

But now doesn't seem to be a good time to invest in exciting and untested new systems of oppression.

We seem poised to access some of the foundations of world (as we currently know it). If we do, then we will face tough questions on who we want to be as a species. Weigh the definition you have in mind and turn it over mentally: is it up to this task?

7.e. Mitigated Hatred

In 2017-2022, I finally experienced the feeling of hatred. At first I was doubtful. I wondered, could I really hate someone without realizing it? Then I noticed a group of political people where just thinking about them and their smug faces made me angry.

Now that I have experienced this feeling, I know that I don't hate anyone for their race. By knowing who I hated, I was able to tell who I didn't.

This hatred has faded somewhat with time and understanding, and who it falls on has narrowed. Now, when I look upon the broader group, rather than focusing on what is there, I find myself focusing on what is missing.

8. Addressing Marlemane's Criticism

This section is going to be broad, and I'm going to cover multiple potential criticisms which are speculative and which Marlemane may not specifically endorse, partly because it brings up topics I want to explain to the general readership, and partly because I don't want to get into an extended back-and-forth to draw out these criticisms.

Marlemane wrote:

They see theirself as a propagandist for their causes, so used to carefully sandwich the reactionary points in between reasonable statements. Guess they’ve decided its time to go mask off.

Consider the program I laid out in section 7.b. What does it do? It continually erodes the material and psychological incentives for ethnic conflict, and thereby continually erodes ethnic tension.

8.a. Reaction

There are many on the left who would like to see an increase in tension, which would crest like a wave and wash over society. In this sense, someone who enhances tensions is a revolutionary, and someone who erodes tensions might be the opposite of a revolutionary – a "reactionary."

Is Mitigated Chaos purely emotional, irrationally holding on to particularities of the past? No. Mitigated Chaos holds a portfolio of bets, carefully considered, with the particularities of different regions, peoples, and individuals contributing to friction, dimensionality, and the resilience of humanity as a whole.

Is Mitigated Chaos purely reactive, merely blindly pushing back against those who want to change the world for the better? No. Mitigated Chaos is reformist, opportunistically seeking improvements while attempting to conserve what matters.

Is Mitigated Chaos "far right," seeking particularist world domination on behalf of some specific ethnic group? No. What about crushing social authoritarianism? Also no.

However, the blog did start off with the tagline, "Revolution is overrated." This blog is pretty clearly against armed Communist revolution, which, given the history of how those tend to turn out, is the sensible liberal position.

8.b. The Empty Space

To someone who would claim that Mitigated Chaos is "reactionary," I would ask, can you see the empty space?

We are born of the world, and so our desires are and have been shaped by the world. We hunger for the food of the world. We smile to see the flowers on the bright green plants under the clear blue sky of the world. We thirst, and on this Earth, nearly everywhere, there is water. And so, we are aligned with the world.

And also, we are not aligned with the world. The alignment occurs on a relatively high level of the environment. Beneath that level, in the specifics of time and place, where human beings live, the world becomes unaligned and drifts away from what we desire. Without planting, the fields will not yield crops. Without maintenance, the wood of a house will become exposed to the elements, and rot. Other creatures, like wolves and bears, contend with us. We use our conscious will to impose the patterns we want on the world.

The world is bountiful and beautiful and fertile. The world is dangerous, scarce, and ugly. That's just the duality of existence.

Sometimes suffering is the result of agentic action. Sometimes it is not.

To overcome suffering which is caused by agentic action, contention, and opposition, one must understand what is caused by agentic action. To overcome suffering which is caused by inaction and emptiness, by the lack or absence of will or materials, one must understand what is caused by inaction and emptiness.

Can you see the strength which is uncultivated?

The farmer stands between you and the field. Can you differentiate between what is done by the farmer and what is not done by the field? Can you tell the difference between oppression and mere suffering?

In order to understand that which is oppressive, one must understand that which is not oppressive. In order to understand when a system is not protecting people, one must be able to understand when the system is protecting people.

To devise systems which insulate people from the harsh edges of reality requires understanding just what it is that we are insulating them from.

8.c. Regarding Group Reputation

I'm going to use a little more bolding here because this subsection is important.

If you actually care about people... If you actually have compassion... If your compassion is real...

...then you must care about something more than just group reputation.

If you actually care about people, then you must care about outcomes for people. If managing group reputation is more important to you than outcomes for people, then the thing that you care about is group reputation, not people.

If women are on average shorter and lighter than men, and someone is devising aircraft ejector seats and refuses to consider this because "being known as shorter than men would be bad for women's group reputation," then the weight sensor in the ejector seat could end up miscalibrated, and women could die.

Believing that women are shorter might seem mean. Designing a faulty ejector seat that is unsafe for women because someone refuses to accept or even consider that they might be shorter is well beyond mean.

I'm not going to tell you that I have an infinite amount of care for other people, because I don't. I'm not going to tell you that I'm entirely against all forms of activity that might get people killed, because I'm not.

But there's such a thing as being wastefully immoral.

8.d. Regarding Criticism

As I once told collapsedsquid:

I can't make any incentive system so complete that it would force anyone to act like a good person all the way through. You have to make a choice.

I have a good grasp of the theory of social and political maneuvering, and I can determine whether someone is using social / tribal, formal, or post-formal moral reasoning.

I also have a fairly good ability to read people.

If you wish to move me, use sincerity.

9. Understanding

Given all of that, where are we now?

9.a. Conservatives

In what will probably not come as a surprise, I don't hate conservatives, people lost in time, pursuing a vision from a world that no longer exists. What is new is not always best, and from a balance of forces perspective, the maintenance of good social norms may depend on people who are not "on trend."

A conservative thinker who clings to his old values sincerely may hold institutions and societies together.

It is my opinion that GhostPalmTechnique is a conservative for a version of the Democratic Party that stopped existing sometime after 2016. There was a battle for the soul of the Democratic party, and he lost.

Holding on to the ideals of an earlier Democratic party is not a bad thing in itself. The Democratic Party of 2008-2014 was in many ways better than the Democratic party of today. It was funnier, smarter, kinder. We poured our hopes and dreams into it.

9.b. Realignment

Like most things, it already contained the seeds of its own destruction. The theory of the "Coalition of the Ascendant" (the language of the Washington Post, New York Times, and The Atlantic), or of the "Emerging Democratic Majority" (the title of the book which projected a 'durable political majority,' the authors of which have now followed with another book titled, "Where Have All the Democrats Gone?"), that demographic change would give the Democrats a long-term political advantage, was boiled down and cooked until many Democrats assumed that demographic change would make them invincible.

One commenter on Twitter/X wrote:

Smart liberal analysts recognize there is no hand of God ensuring constant progression to the left. John Judis and Ruy Teixeira premised continued Democratic electoral success powered by demographics on “progressive centrism” by which they meant liberalism constrained by reality.

Holding to a theory that Democrats would no longer have to compromise has radicalized Democrats, as they keep discovering, to their horror, that Republicans can change and remain electorally viable.

They are shocked and horrified by the change. How is it that Republicans could become more economically populist? How is it that they could respond to theories of demographic change with their own theories of demographic change? It seems so unfair! Weren't Republicans supposed to remain a static force that could easily be out-maneuvered?

I spent years trying to explain to Democrats that Republicans could change. Broadly, there were two implicit theories as to why Republicans couldn't.

The first theory is what I call the "bitter core" theory. It holds that there is a bitter core ("bitter clingers," "deplorables," "racists," "xenophobes") of Republican voters who despise minorities, women, etc, and are not willing to compromise. In this theory, a Republican candidate cannot possibly come up with a message that will convince both minorities and the bitter core to vote for him at the same time. The coalition is just not sufficiently coherent.

The bitter core theory has several problems:

Republicans are not actually obligated to be racist for its own sake.

Republican primary voters never signed a contract saying that they weren't allowed to elect unconventional Republicans if conventional Republicans either seemed unlikely to win or were discredited (such as by the Iraq War).

An electoral advantage is not the same thing as a guaranteed victory. A lack of discipline among the party's supporters is something that is difficult to precisely control.

The second theory is just a very common kind of political behavior, which is treating a group of perceived political opponents as a monolithic blob.

In practical terms, politicians cannot act alone, so politics is about power coalitions. These coalitions set limits on a politician's feasible behavior, as the different members of the power coalition will have differing opinions, and the politician needs to keep a power majority on-side in order to act.

Treating Republicans as a monolith leads to several problematic ideas:

The idea that Republicans are always already maximally escalated, so it's impossible for them to escalate further in response to actions by Democrats. (Maximum escalation would look like a hot civil war, not present conditions.)

The idea that Republicans are not currently maximally escalated, but will always escalate at every opportunity regardless of Democratic behavior. (Shouldn't they have escalated more by now, then?)

The idea that Republicans are all uniformly evil, so Republicans responding to a Democratic escalation is just Republicans being evil and has nothing to do with the nature of the escalation itself.

These three approaches tend to treat Republicans as a fixed point rather than a dynamic and ever-shifting force, and as a side effect, tend to treat them as fixed in time.

I tend to take a very different approach to politics (which is why Marlemane called me a "propagandist").

There will always be disagreement within society, and there will always be political teams in opposition to each other, although the composition of the coalitions, and their associated ideologies, will change over time. It's better to think in terms of the individual issues, rather than the teams.

Winning doesn't look like crushing the other team. Winning looks like one's positions on the important issues becoming the new normal. I have won when my position is "not political," and my opponents are marginal.

From this perspective, it is better to achieve a favorable compromise with supermajority support than it is to try for a "perfect" outcome and fail. This does not mean that the world cannot improve; it just means that improving the world takes actual work and consideration. Likewise, it doesn't mean never using force; a large military force actively engaged in warfare generally requires an opposing military force to stop.

Winning looks like getting a majority, or near-majority, within each party. The parties are just vehicles for achieving the desired political outcome.

If people become healthy, happy, and strong, if they think through policy and weigh options and seek truth, if they can learn to balance compassion and pragmatism, then the worst policies will be excluded from consideration. If the worst policies are excluded from consideration, then I have won.

By living at peace in the 21st century, with its material abundance and high life expectancy, I am already winning.

9.c. Perspectives on Principles

There was quite a bit of coverage of an "open letter by scientists" in 2020 supporting the national protests over the death of George Floyd. At the time, a writer for Time Magazine wrote:

“We support [racial-justice protests] as vital to the national public health and to the threatened health specifically of Black people in the United States. We can show that support by facilitating safest protesting practices without detracting from demonstrators’ ability to gather and demand change,” the letter reads. “This should not be confused with a permissive stance on all gatherings, particularly protests against stay-home orders.”

In theory, prohibitions against church gatherings infringed on the human rights of religious Americans to freely practice their religion. Likewise, attempts to prohibit protests against stay-at-home orders (which, say, religious people might attend) would be an infringement against the free speech rights of those protesters.

Should the letter's authors have noticed that they were violating general principles?

If you view the concept of free speech as a kind of peace treaty, where we agreed to stop shooting each other on the condition that we be allowed to express our thoughts freely and resolve our disputes through intellectual competition instead of violence, then allowing left-progressive protests while prohibiting right-conservative protests is a betrayal of that deal. It's a treaty violation, and a selfish one, daring the opponent to do something much more aggressive to restore their rights.

If you're someone who thinks in these terms, a lot of 2014-2024 left-progressivism isn't just a violation of agreements; it's also confusing.

For example, when I was young, in public school, we covered the existence of various historical atrocities or historically oppressive laws on the basis of race or ethnicity. From this, I inferred that people have, historically, had a tendency to go a bit crazy about race/ethnicity to the point that they'll do things like murder (including, potentially, very large-scale murder).

From this perspective, left-progressives trying to promote "racial consciousness," especially in the racial majority of the country, came off as insane. It isn't just that doing so was selfish. It was also likely to be counter-productive. If people were "racially awakened," it was possible that they might embrace conscious racism, rather than reject it.

I've come around to the view that the median left-progressive doesn't think in these terms.

The question is whether someone views free speech as a concept that exists independently of tribe or hierarchy, or whether they view it as a social norm.

In the first view, the concept of free speech is based on something I call "soft game theory."

Consider markets. Markets require certain conditions in order to run, such as participants not constantly lying to each other, buyers being able to judge the quality of products, and so on. However, markets also reflect regularities in the nature of the economic planning problem (such as the distribution of information), and thus work anywhere those underlying conditions hold.

From this perspective, free speech is viewed in terms of its mechanical properties. Someone may believe that open discussion leads to an improvement in knowledge, that freedom to speak does not guarantee that the majority will have the correct opinion but allows a minority to maintain the truth, or that allowing politicians and committees to determine allowed speech will lead to those politicians prohibiting speech that is inconvenient to them (rather than merely speech which is more directly dangerous such as bomb-making instructions). (I discuss my version of this in my "Free Speech" post.)

Someone can define this in formal terms, as the field of economics defines things, but one can also think about this intuitively.

In the second view, freedom of speech is a social rule that either comes down from within-group authorities, represents the group consensus, or both.

Social rules are much more particularistic than something like economics. Something like fashion, or colors for differentiating two different groups, may be a social rule. Social rules usually have some point to them, but it's assumed that the leaders of the group, or of society, have wisely decided them.

You're not actually supposed to examine the social rules independently. It would be considered strange if you did, such as if you started wearing all of your T-shirts inside out.

From this perspective, if the leadership (such as socially trusted information sources such as the New York Times, or 'experts,' or celebrities, or a highly socially adept person in one's network) say that it's time to replace or limit the social rule of free speech, such as to "police hate speech," then it's time to replace or limit the social rule.

Alternatively, it will be taken as "obvious" that a beneficial social rule does not apply to outsiders, or that if there is a positive quality insiders must have it by definition, while if there is a negative quality, outsiders have it.

As such, it is likely that many of the people who supported allowing one set of protests, while prohibiting another set, don't think that they "hypocritically violated the general principle of free speech." They think that they "courageously stuck to the principle of upholding social rules provided by rightful experts."

Someone who takes the first view generally thinks in terms of procedures, actions, and models.

Someone who takes the second view generally thinks in terms of people, prestige, and titles.

Each discursive pattern has both a surface style (such as "I'm a cold badass who only thinks in terms of numbers; this is the cold reality you must accept,") and a substantive basis from which that style derives its meaning or power (such as actually studying the numbers and actually ignoring social taboos to study them).

People who take the second view tend to think excessively in terms of public relations, reputation, and the manipulation of words to alter reality. They seem to have difficulty understanding the mechanical basis from which the prestige of the titles is derived.

This is not to say that there is no merit in thinking of social effects.

Rather, blindly following social consensus is one of those strategies that works until it doesn't. Someone has to do the work of setting a sound social consensus, and that will almost necessarily entail violating the existing social consensus in some way. Likewise, there will be situations that differ from social consensus in the particulars, even if the social consensus is reasonable in more general situations.

It is good to be socially adept, and to understand what the current social consensus is. It is also good to have a brain, a spine, and a sound sense of personal judgment, and thus know when to follow that social consensus, and when to violate it.

At some point, obsessively adhering to the local social consensus stops being socially adept.

The people who supported what the New York Times called the "racial reckoning," while opposing other protests, created opposition that they did not need to create, and radicalized opponents that they did not need to radicalize. In that climate, with op-eds with titles like, "Yes, We Mean Literally Abolish the Police," ordinary police reform got pushed out of the national conversation. (To pick a case reported on in 2024, there were reports that police threw a flashbang grenade into a room with a baby on a ventilator... at the wrong house. I have no doubt that there is room for improvement.)

Some Tumblr users have already noticed a related phenomenon, in which people don't learn e.g. not to bully people, they just learn "a list of socially acceptable targets." I also find that annoying, though beyond that, it may lead some Tumblr users into despair.

9.d. The Crystal Prison

They say the first 10,000 years in the crystals are the longest I hope your little Trump doubter post was worth it, pal!

- right-wing Twitter/X user MysteryGrove, October 13, 2024

Shortly before the 2024 U.S. Presidential election, the intellectual vanguard of the online right began saying that they would trap their opponents in "the crystals" for a thousand years. This was applied to both 'the libtards,' and to perceived Trump doubters, such as Nick Fuentes and his followers. (One leader described Fuentes as "memetic pollution.")

This was presumably annoying to their opponents, but (at least in the United States), in no way legally actionable. (Some uses were a bit more whimsical.)

After the election, as blocklist-enthusiast left-progressives departed from Twitter/X, grumpy anti-institutional centrist TheAgeofShoddy asked:

So was the crystal everyone was talking about just bluesky?

When I write science fiction posts on this blog, my goal is often to broaden people's perceptions or to satirize. (One blog, now vanished, said that I was among the best satirists of the present day.)

In the science fiction post "Re: The Chillers Discourse," it's implied that "sixth-generation warfare" is "incomprehensible." In the "Thermal Wars TCG" post, "Ideological Recomposition" is a form of sixth-generation warfare attack that adds one point of ideological tension to an opponent's coalition, while subtracting one point of tension from one's own.

The tags from one post from 2020 say that, "Ninth-generation warfare sees all acts as existing on a spectrum of political violence. Most acts of ninth-generation warfare consist of extreme pranks."

These are all references to "fifth generation warfare," which Wikipedia describes as "non-kinetic military action" in a "war of 'information and perception'". This theory is apparently disputed, which is unsurprising: at some point a "war of perception" strains the definition of what can be called "war."

Let us turn the situation on its head.

Let us imagine an act of dimensional warfare.

Suppose that Donald Trump possesses an extremely powerful weapon, a crystal staff. (It's magic, just so you know.)

The weapon reduces the amount of meaning in text. It might make longer texts shorter, or reduce the amount of meaning per word, requiring bombastic rhetoric ("we must destroy all car-owners") to make ordinary claims ("we should build more bike lanes"). (I discuss word-meaning reduction in section 'c - The Problem 2' of "The Low Friction Moment".)

If there are two or more opposing opinions on some issue in a left-leaning environment, the weapon crushes holders of the minority opinions in order to force a majority-opinion consensus, regardless of whether the majority opinion makes sense in a particular context, or makes sense in general.

By crunching multiple opinions down into one opinion, this weapon would reduce the number of bits needed to represent left-wing views, and fix left-wing dogma in a particular configuration.

U.S. left-wing ideology would become rigid and inflexible, like a crystal, and anyone who disagreed with this process itself would also be driven out.

A certain amount of disagreement and tension within a political coalition is normal and healthy. Different people have different expertise, different interests, and different life experiences. There is a trade-off between diversity and uniformity.

We can think of it in terms of steps of a process. 1) We want to be broad in what we initially think about. 2) We then want to narrow that down through study, experiment, theory, and consideration. 3) In policy, we want the law to weigh multiple considerations or situations. 4) Then, we want the enforcement of law to be relatively uniform.

The virtue of flexibility is in the cycle between 1 and 2, or 2 and 3. If we have multiple competing strains of thinking or schools of thought, then if a policy fails (3), we can roll back to the previous step (2) and try again with an alternative that's still reasonably acceptable to the members of our coalition. In political arguments, if we have a diversity of ideas that we consider initially (1), then if an opponent comes at us with a proposal that we don't have a better or more compelling alternative to (2), we can roll back to the previous step (1) and develop a new alternative (2).

With one wave of his crystal staff, Donald Trump collapses all the alternative policies in step 2 down to one, and all the alternative ideas in step 1 down to one. With this act, he freezes the Democrats in place.

Most people don't think of it in this way, but coming up with compelling political ideas is actually work, and it takes talented people to do it.

In general, the U.S. right have been suffering from a shortage of talented people. This isn't surprising, as the Iraq War was quite discrediting for the Republican party, and Democrats were a more moderate party in 2008, relative to today. Even if someone were against immigration, that was more acceptable among Democrats back then, and if someone were worried about American power, well, the wars in the Middle East weren't great for that.

As a result, Trump doesn't have the same number of intellectual staff that the Democrats do, but with the magic of the crystal staff, he doesn't need to.

It doesn't matter if the Democrats have more guys with a higher IQ if they're not allowed to use that IQ. It doesn't matter if Democrats have more guys with expertise if they're not allowed to have their own opinions.

Trump managed to recruit an intellectual vanguard. It may not be the biggest, or the smartest, but it exists. If they provide him with 3 options, and Democrats would have had 5 options, but were limited down to 1 option, then Trump and his Republicans still have the mobility advantage. Trump can let the frozen Democrats take an unpopular position, and then just keep hammering them over and over again, because they cannot dodge.

This is part of what having multiple options is for – dodging.

There is a second problem.

There are political positions that make sense in isolation, but which don't make sense or have bad effects when put together. (People may even extrapolate backwards from the combination and assume that someone advancing that combo is a lunatic who is causing problems on purpose.)

For example, opposing construction of housing and public transport infrastructure on environmental grounds or imposing many rounds of community review might make sense if the size of the population in an area is fixed. Based on common liberal or left-leaning assumptions, mass immigration may make sense.

The combination of refusing to build new housing, while bringing in millions of people who will need new housing when they arrive, is ridiculous. It will put tremendous stress on renters, and cause a homelessness crisis under conditions where one would not otherwise exist.

Someone has to come through and harmonize the policy. Someone has to act like an actual leader, not one of a dozen competing interest groups, look at the set of policies proposed, and change the policy set if it does not make sense.

To do this, they need options. If a politician comes in and the acceptable range of opinion and policy give them five options each for the amount of immigration and the amount of construction, then they can likely exercise a little will and political talent and pick two that are in balance.

With a wave of the crystal staff, that goes away.

With only one option for each issue, there's no requirement that the ideas match up. The network of ideas could end up contrary to sound government, and commit Democratic strongholds to bad (and embarrassing) outcomes. This would make it easier for Donald Trump to campaign, because he wouldn't have to work as hard to offer, or at least sell, a perceived higher quality of government.

If such a weapon existed, activating it would be a war crime. By solidifying the position of the opposing faction, it would prevent opposing leaders from being able to negotiate peace. This would lead to the crushing of entire populations.

At first, being subject to such a weapon would make its victims feel powerful, as the unity of their coalition increases and they triumph over internal competitors. The downside of the formation becoming inflexible and brittle would not be fully noticed until later.

Of course, there is no such weapon.

Members of the Democratic coalition did this to themselves. I'm sure it seemed like a good idea at the time. The media landscape from 2014 to 2024 was quite tumultuous for a number of reasons.

Attempting to hide or deny Biden's cognitive decline is merely the most obvious manifestation. The entire Biden presidency was characterized by a lack of agency from the top, and an inability or unwillingness to discipline elements of the Democratic coalition and thereby harmonize policy. This probably also seemed like a good at the time.

We could imagine a situation that is even worse.

The weapon could suppress someone's ability to differentiate between suffering caused by oppression and suffering caused by the environment or a lack of human action.

This could lead to all sorts of nasty consequences.

10. Additional Concepts

Before continuing, I'd like to cover a few additional concepts.

10.a. Layer-Based Analysis

Because the world is so vast and therefore difficult to describe (as per the earlier discussion of dimensionality and the Dao), it can be difficult to get a grasp on it.

One technique I'm particularly fond of is to separate the system or situation to be described into multiple layers (at differing bands of abstraction). Once I have divided the system into layers, I can model out the interactions within each layer, where characteristics are similar. Once I have modeled the interactions within each layer, I can then go back and model the connections or interactions between the layers.

I'll provide an example.

In real life, scale is continuous. There is no sharply-defined boundary between the scale of a single molecule in a cell in your kidney and that of your entire body. However, the effects of a single molecule are unlikely to travel very far, due to the typical amount of energy and mass involved. Effects generally won't be noticed at the scale of the body until many molecules are involved.

Let's sketch out a very simple model of disease. (I am not saying that this is how disease actually works; this is just a demonstration of motion across and between the layers.) There will be two layers: the human-scale layer, and the micro-scale layer.

◇ micro bacteria enter the body and begin multiplying

→ micro the immune system detects the presence of hostile bacteria and sends chemical signals that raise the body temperature

micro ↑ human the human notices a fever and realizes he is sick

→ human the human goes to a doctor and receives antibiotic medication

→ human the human swallows a pill containing antibiotics

human ↓ micro the anti-bacterial chemicals diffuse in the bloodstream

→ micro bacteria die and their population falls

→ micro the immune system no longer detects the bacteria and stops signalling to raise the body temperature; the fever clears

micro ↑ human the fever clears and the human notices he is no longer sick

This method can also be used to divide an object up into characteristics or qualities that aren't on a single fixed spectrum.

Consider a well-crafted prop steel I-beam on a set in Hollywood, made out of styrofoam.

visual - it looks just like a real I-beam

mass - it doesn't weigh anything like a real I-beam, and could blow away in a breeze

sturdiness - it would be fairly easy to snap in half compared to a real I-beam

So, is our fake I-beam like a real I-beam, or not?

The answer is, it depends on how we're interacting with it. If we're just viewing it, then it's like a real I-beam. If we're trying to hold stuff up with it, then it isn't. It depends on who is asking, and why.

Our fake I-beam has mechanical qualities that exist independently of what we think about it. It is mostly not like an I-beam, for the purposes that most people would have in mind when they use the term "I-beam."

Therefore, it we want to sell it, we should split the concept up. Calling it a "prop I-beam" would get the appropriate differences across quite well.

10.b. Managing Dimensionality

Imagine a point on a flat line. Easy, right? Now, imagine a line on a two-dimensional graph. It's straightforward. You've probably seen a hundreds or even thousands of line graphs in your life. Now, imagine a point in a three-dimensional space, with an X, Y, and Z axis. That's a bit more difficult if you're not used to it. (If you'd like to get more used to it, I'd recommend the 1999 video game Homeworld.) Can you imagine a point in a four-dimensional space?

The more complicated a model is, the more difficult it is to process. This is true for humans, but it's also true for computers. (This is why computer programmers are interested in Big-O notation, as mentioned in section 3.c.)

The more dimensions there are, and the broader or more detailed each dimension, the more possible combinations of outcomes must be considered. If there are five models of car by a manufacturer, and each one comes in five colors, then there are 25 options to consider. If each model has five trim levels, then there are 125 options to consider.

It's a simple result of multiplication. This is why, in section 3.a., I wrote:

All human models must reduce the number of characteristics (or dimensions) considered, and reduce the number of options per dimension (scope of a dimension) considered.

Suppose that it takes me 1 minute to evaluate how much I like the color of a car, 1 minute to evaluate how much I like the model, and 3 minutes to evaluate how much I like the trim. If I evaluate each car in isolation, then it will take me 625 minutes, or about 10½ hours, to evaluate all the cars. [1]

Let's suppose that the 5 colors are the same for all the cars. In this case, I may be able to decouple the color from the other characteristics and consider it separately. As such, I spend 5 minutes picking 1 favorite out of the 5 colors. This converts what had been multiplication into addition: (5 options × 1 minute per option) + (25 options × 4 minutes per option), for 105 minutes.

Perhaps I decide that I have plenty of cash, the minimum equipment provided to all cars by this manufacturer is fine, and I want the car today. As such, I'll take any new car on the lot as long as I like the model and color. This removes trim level from consideration entirely, and the whole process then takes 10 minutes (5 colors + 5 models).

The art of analysis is a form of labor.

When is it appropriate to split a concept up, and thereby add more information to the model, while also requiring more effort to use the model? When is it appropriate to merge two concepts together, and thereby remove information from model, and make the results faster to estimate? What information is irrelevant to the current question and can be safely ignored? What information provides the best predictive power for our question, and is essential? How much predictive power can we afford with our limited amount of time or resources?

These are all decisions which have to be made, and which require judgment. Even thinking about how much research to do takes time and energy.

10.c. Delegation

Suppose you are issued the order, "Build a new locomotive." (Let's also say that you are provided a budget.)

Now of course, if you are one person, you've completed calculus, and you have 30 years to do it, you could probably learn all the necessary information yourself from experimentation and books, smelt some iron and build a primitive locomotive of your own design. There was once a time when there were no locomotives, so someone had to be the first guy to devise one.

However, suppose you want to build it faster.

We can view building the locomotive as a problem with its own dimensions. If the problem is too large for one person, then we can divide that problem up into smaller parts and split the problem up among multiple people.

We can hire a mechanical engineer to design the locomotive. We can hire a machinist to construct it.

The two of them will need to be able to communicate about the design. That's a form of overhead that didn't exist when the job was done by just one person. We will likely need to provide them with an email system.

Likewise, our machinist and engineer may argue with each other about the design. Someone will have to either bring them to an agreement, or make a decision about what the best approach is. (The flow of information and decision-making creates a natural hierarchy.)

The engineer and machinist will need to work together closely to design and construct the locomotive. For other parts of the problem, we can delegate more loosely. For example, we can order steel from a supplier, and rent out a shop and equipment from a third party.

Describing my abstract theory of capital, I once wrote:

Capital is a low-dimensionality construct [...] Labor is mostly about reducing the context of a problem until it is possible to apply capital, and then applying that capital.

A worker may place a sheet of metal into a metal stamping press, and then press down, reshaping the metal sheet. The worker removes other considerations from the situation, like other random objects being in front of the stamp press, so that the design of the stamp press does not have to consider them, and can be simple.

As the director of this project, your responsibility will be to reduce the context of a problem until it's just a job. You will take care of ordering the steel so that the machinist does not have to know about your dispute with the mill over the delivery date. You will procure the shop so that your machinist doesn't have to think about when the rent payments are due. That way, he only has to argue with the engineer about the size of the boiler, which is part of his actual domain of expertise.

Organizing and coordinating production is work. Organizing and coordinating production has costs.

10.d. Formations

Our locomotive project has multiple members. These members operate in synchronized action. This synchronized action produces a higher-dimensionality or more complex force (than the actions of any one member). This higher-dimensionality force produces a locomotive.

If the engineer decides she doesn't want to design a locomotive and instead wanders off to design a car, then the actions of the formation are no longer synchronized, the force will become incoherent, and a locomotive will not be produced.

Talent and skill matter.

For the locomotive to be produced, each part of the process must go correctly. The design must be of adequate quality. The steel must arrive in good condition and on-time. The design must be built out according to specifications.

If we hire someone who is not able to do mechanical engineering, then he may not be able to provide an adequate design to the machinist, who will be left to try to hack together something himself.

Social skill and temperament matter.

If the engineer and machinist both despise each other, then they may refuse to work together or even sabotage the project.

If the engineer and machinist start arguing, and you as the director are not able to help them to calm down and work together, then a gradual build-up of tension from arguments may occur until the engineer leaves halfway through the project in a huff, you have to hire another engineer, and the locomotive may not delivered before the deadline.

If you as the director are not able to judge potential hires, you may hire the wrong guy entirely, have to fire him, go through the trouble of hiring someone else, and miss the deadline.

Decision-making authority matters.

Reducing the scale of a problem by dividing it up also means dividing up decision-making. If one person designs and builds the locomotive, then he makes all the decisions by himself. If he contracts out the engineering of the design, then he necessarily will not be making all of the decisions on the design as though he did it himself, because if he did, it would take the same amount of time and effort as doing it himself, which would defeat the purpose of hiring someone else.

However, the engineer and the machinist might have some irreconcilable dispute about whether some design element is worth the risk, or whether it's better to use one design that eases maintenance or another that increases horsepower. If they have to have a 6-month vim vs emacs style argument about this, then actually building the locomotive may grind to a halt. [2] Someone has to have the authority to pick one option and make both the engineer and the machinist stick with it.

As the project director, your employees have more fine-grained information, while you have more general information about the context of the project. Likewise, your employees naturally make more fine-grained decisions about the project, while you make more general ones.

If you have one engineer and one machinist, then each one can write a daily report in 30 minutes, to be read by the other in 5 minutes. (Writing emails is work, too.)

If you have ten engineers and ten machinists, then each one will be spending 95 minutes reading reports each day. The organization then tops out at 90 employees, at which point no work is getting done.

(The cost of additional communication cutting into working time is a well-known matter in the field of project management for software development, most famously discussed in 1975's The Mythical Man Month.)

In a large project with many members, communications overhead requires further division of the workforce into teams, requiring multiple levels of management. Communications overhead and the ability of managers to manage impose limitations on organization size, capability, and coherence.

In my post on abstract capital, in discussing the original poster's point about degrowth, I also wrote:

Holding the complexity constant, the smaller the population, the greater the share of the production problem that each worker must understand.

An organization with more skilled, talented, experienced, or simply smarter employees may be smaller, and thus have lower communications overhead. Wiser and more intelligent managers may also be able to keep a larger team's work more coherent.

10.e. Belgium

Let's practice some analysis using a scenario from another post of mine.

Let’s say we’re France and we’re having ethnic tension with Belgium over who has the best cheese. There is an army group within the Belgian army who were humiliated during a riot at an international cheese festival that left 7 dead and 145 injured.

The army group have been conducting unscheduled “military exercises” very close to the French border. We want them to stop that. Ideally, the leadership of the group would be dismissed.

As such, we attempt to get leverage on the Belgian government by threatening to ban the export of baguettes and croissants into Belgium. This will negatively impact Belgian civilians, who buy them regularly for some reason.

A fellow French government officer turns to you and says, "Damn, those Belgians are sure up to some bullshit, stomping around on the border and threatening us Frenchmen. We better get revenge by cutting off their supply of baguettes!"

Well, not so fast.

First, divide the concept of "Belgium" vertically, into two layers.

National - The shared interests of "Belgium" as a whole, as if "Belgium" were a rational actor.

Sub-National - Groups (including agencies) or factions within Belgium.

When we say we need to "get revenge on Belgium," this only makes sense if "Belgium" is coherent. Is the trampling along the border an intentional act of "Belgium"?

Let's divide the sub-national layer horizontally.

Rogue Army Unit - The guys stomping around on the border.

Legislature - The official civilian government of Belgium.

Belgian Citizens - Belgians not affiliated with the military.

Now, we'll map the relations between these factions. The Belgian citizens pay the legislature and elect the legislators. The legislature pay the Belgian army, including the rogue army unit, and presumably appoint members of the staff. (I don't know how armies work in Europe.)

So the baguette plan looks like this:

French government (us) bans export of baguettes to Belgium → Belgian citizens get angry at loss of baguettes, threaten to vote out legislators → To avoid being voted out, legislators order the rogue army unit to stand down → If the unit do not stand down, the legislature order that the unit members be fired and arrested

If Belgium is operating as a relatively coherent formation, the legislators will have internalized the logic of this sequence. Thus, by our threatening to ban export of baguettes, they will respond to solve the problem before our ban even goes into place, and no ban will occur.

A few ways this could go wrong...

The legislature are in fear of the rogue army unit, which is why they didn't bring them back under control already. They deflect the voters' anger at the loss of baguettes back at us, the French, increasing ethnic tension and Belgian citizen support for the rogue unit's actions.

The legislature are incompetent and are not capable of responding to signals from voters. There is also no chance a competent party will get voted in. No matter how much pressure we put on Belgian civilians, there is no machine to convert this into influence on the rogue army unit.

Belgian citizens don't view it as their responsibility to influence politicians to solve the rogue army unit problem – from their perspective, that's the French government's problem. They get angry at us for cutting off the baguette supply.

Another way to think of this is that international politics are really about domestic politics. What we want is to alter the balance of power of political factions within Belgium.

The rogue army unit are most likely justifying their actions with, "Our hearts are blazing with patriotism after the deaths of our comrades in that cheese riot!" This makes it politically more difficult for the legislature to oppose them. Our threat to cut off the baguette supply is actually a tool to provide the Belgian legislature with political leverage over the army unit, by leveraging the hypothetical anger of the Belgian citizen voters. ("I'm so sorry, but if this ban goes through, the voters will be furious...")

However, this suggests that there may be alternatives to the baguette embargo to achieve the same goal.

11. The Labor of Peace

Neither war nor peace are the default state of humanity. Rather, we are a species who engage in cooperative competition.

War and peace are options. Peace is a choice.

11.a. A Peace Tree Search

We can view each person as having multiple options for what to do at any moment, like walk, call someone, do a back-flip, or scroll Twitter.

We can view these options as requiring different amounts of effort. Doing a back-flip, for example, would probably take a lot more effort for most people than scrolling twitter would. There may be some choices that are not options simply because they are too difficult.

In considering the combination of how people think with how difficult the actions are, we can say there there is a different likelihood or probability for each action. This isn't to say that free will doesn't exist; it's just a way of thinking about things.

While taking the action, someone moves forward in time. They then (usually) have the opportunity to take another action. So for example, someone might do a back-flip, decide that was a bad idea, and then decide to do something easier, like scroll Twitter.

We can imagine all the paths someone might take, with all of their options, stretching out before them in time. Some options may foreclose other options.

There may be potential paths that someone can't see. They might go to the supermarket and find a product for sale that they didn't know existed. There may be potential paths that someone thought they could go down, but can't. For example, there may be a gas leak which prevents them from going to the supermarket that day as planned, as emergency crews secure the scene.

When two people interact, each one may choose to do so in a neutral way, a hostile way, or a beneficial way. For example, when two men walk past each other, one might choose to rob the other (a hostile act), nod his head in greeting (a positive act), or just keep walking and ignore the other man (a neutral act).

Peace is the outcome when both men choose the neutral or beneficial act, while conflict is when at least one chooses the hostile act. There can be different levels of severity of conflict, from outright violence to a mere insult. Conflict is also not always the worst outcome. (Sometimes conflict in one layer, such as social conflict, can prevent conflict in another layer, such as violent conflict.)

When the two men approach each other, each one does so with some level of uncertainty. Because each one has freedom to act, it is always possible, even if very improbable, that the other man might engage in a hostile act.

Each man has the choice between peace and conflict.

Therefore, to create peace, each man must be willing to take on a little risk, and act on faith. That doesn't mean he should never use violence; rather, violence is something that should only be used sparingly and with great care.

People also make mistakes, so it is appropriate to practice forgiveness. Again, this does not mean practicing unlimited forgiveness. To quote the version of The Analects from Project Gutenberg:

Tsai Wo asked, saying, 'A benevolent man, though it be told him,— 'There is a man in the well' will go in after him, I suppose.'

Confucius said, 'Why should he do so?' A superior man may be made to go to the well, but he cannot be made to go down into it. He may be imposed upon, but he cannot be fooled.'

With a greater level of intelligence, someone can imagine more possible actions and estimate more possible consequences. With greater wisdom, someone can more accurately judge people's character and their likely behavior.

Thus, with both together, someone can search farther down the paths to find a mutually beneficial – or at least mutually neutral – deal to offer.

The labor of peace involves altering the relative balance of probability, and of effort, of the different paths of potential future actions, so as to make peace easier and more likely, and conflict more difficult and less likely.

This may, in itself, involve conflict. Arresting and imprisoning a man who commits armed robbery will involve being in conflict with him, but imprisoning him will reduce the expectation of other citizens that they will be robbed and that, therefore, they may have to act quickly with violence to protect themselves.

In theory, everything I just said above in this subsection should be obvious. Then again, there are a lot of things I thought were obvious in the past that other people apparently didn't think were obvious.

11.b. Peace Through Strength

Creating a peace where it did not exist, enforcing a peace that exists, and maintaining a peace that currently exists, are all tasks that may be as easy as saying, "Hello," or more difficult than fighting and winning a war.

Physical strength can help. It provides you with options. With strength, you can stand tall and resist intimidation. You will have greater freedom to be merciful, and will be able to save others who cannot save themselves.

To truly create peace, intellectual strength is also necessary. While something like IQ is very difficult to change, it is possible to learn more about the ways of the world and the nature of human beings, to practice thinking and analyzing, and thereby anticipate both events and human actions.

Social strength is essential. The ability to understand others, to know their motives and what is important to them, is necessary to come up with deals that they will like. The ability to charm others, whether by inspiring them or by being disarming and inviting, can melt a social situation that was on the verge of becoming a fight, and help the parties reach an agreement.

There is also moral strength. You will have to consider the nature of morality and take right actions in order to develop moral authority and be able to cooperate with others. You will have to take on risk and act on faith without knowing for sure whether others will cooperate. Others, who may be more self-interested or thinking through a more tribal lens, may oppose you, and you will have to have the courage to stand up to them, or the endurance to quietly go around them. You will also have to practice humility (1, 3), in case you, yourself, have misjudged the world.

To create peace and maintain justice, you will need to learn to see the empty space (8.b.), and separate out what suffering is caused by agentic action, and what suffering is caused by the lack of action. You will need to be able to tell the difference between when the correct response is to teach others, train them, and help them become strong, and when the correct response is to punish the wicked or remove them from power (6.c.).

You will need to be ideologically flexible (9.d.), not dogmatically set to any fixed possibility (4), and explore options that will find a path to peace. If there do not appear to be any such options, then you will have to step back (3.c.) and see if you can create one.

You will also have to be able to tell the difference between what you can change, and what you cannot change (3.c.), so that you do not put yourself in conflict with others when you do not need to be (6.e.).

You may have to train and coordinate others in a large group (10.d.). If there is no leadership that can broker a deal (10.e.) (that is, negotiate a binding agreement), then you may have to create it. If no one else steps up, you may have to become the leader that can make and enforce the deal yourself.

Thus, to make, enforce, and maintain peace, you will have to have a high dimensionality (3.a., 6.g.) in order to have options, be aware of the situation, and maintain maneuverability. In other words, you must be like a more complete human being, flexible yet strong, and resist reduction and subjugation. This does not mean that you should never invest in a position and thereby reduce your future options, but rather that there are trade-offs and you should maintain an open mind.

Seeking peace may involve using sophisticated strategies to erode the desire for conflict (7.b.), and it may involve being misunderstood by others (7.c.) who cannot see – and therefore cannot verify – your inner motivations.

This is why I call it the labor of peace. While there are often regularities, the world and its conditions are always changing, and therefore peace requires ongoing effort.

That is, peace is a choice, and it is one that you will have to keep making. This isn't the worst thing in the world; it is because peace is a choice which comes up again and again that peace can once again become a choice during conflict.

Because requires strength, wisdom, and effort, including courage, the man of peace may be as worthy as the man of war.

11.c. The Root of Politics

Material conditions are foundational. Without material conditions, there is not existence, and without existence, there cannot be politics. However, material conditions are not politics. The grass in the field is not itself political. Politics arises from competing value judgments, often about material conditions. Material conditions constrain political options.

Consider the following four $100,000 options. Which one is best?

A $100,000 sports car.

A $100,000 European vacation.

Receiving $100,000 to have another child.

$100,000 in compensation to enter a career which pays $100,000 less in lifetime income, but which is highly prestigious.

The answer is that it's subjective.

Someone may be uninterested in cars, uninterested in Europe, uninterested in children, or uninterested in career prestige. People have preferences. People believe things. They have emotions, and they also have differing access to information. They may behave rationally, irrationally, or in ways that seem irrational if you don't understand their motives. They are not defined solely by their relationship to capital.

Which arguments someone finds compelling depends on their personality, their beliefs, and their life experiences – their individual psychology.

The personal experience that someone has depends on multiple factors. Partly, it's the result of factors in the external world over which they have little to no control (such as being hit by a drunk driver). Partly, it's the result of choices they make (such as going to a college and picking a major that has many other students similar to them). Partly, it's how they interpret real-life events that would be interpreted differently by someone else (influenced by their personality and prior beliefs). Each person has a different life, and a different way of viewing life. They are going to have different opinions.

11.d. The Coffee Shop

Suppose that there is a coffee shop somewhere in America.

The shop's budget looks like this:

Rent: $2,000 / month

Utilities & Misc Expenses: $500 / month

Coffee Materials: $1.50 / cup

Coffee Selling Price: $4 / cup

Owner Wage: $12 / hour

Employee Wages: $12 / hour

The owner works 40 hours per week. He has two employees – one works 20 hours per week, and the other works 16 hours per week. The shop is open 8 hours a day, 7 days a week (56 hours). (We'll say that one month is a simple four weeks, to make the math easier.)

The shop sells an average of 12 coffees an hour during the 56 hours it is open each week. Thus, the shop's monthly finances look like this:

Revenue: $10,752 / month

Monthly Expenses: $2,500 / month

Materials Expenses: $4,032 / month

Wages: $1,920 (Owner) + $960 (Employee 1) + $768 (Employee 2) = $3,648 / month

Total Expenses: $10,180 / month

Profit: $572 / month

In real life, I'm pretty sure that non-wage employee expenses (such as payments to Federal insurance schemes) are higher, per-cup coffee materials costs are lower, and so on, but you get the idea. $4 is a reasonable price for a cup of coffee on the go, and 12 cups an hour is an average of one cup every 5 minutes.

The owner is making about a 5% profit margin overall. This is actually well within the range of typical net profit margins for most industrial sectors.

Suppose that one barista notes that she regularly sells 12 cups of coffee per hour while she is scheduled. She becomes upset, because the first 3 pay for her wages, while the additional 9 bring in an additional $36 for the owner! She has a sense that the materials cost a lot less than they're selling each cup for.

Many people are not actually familiar with business. They haven't tried to run a business. They haven't taken a course on it. Many business owners simply may not think to explain what everything costs, explain losses from discarded or damaged inventory from maintenance, and so on. They might also refuse to explain, on the grounds that such information is useful to competitors. (They would also prefer not to discuss wages, to prevent workers from negotiating up their wages.)

As employee and shop owner, the owner and barista have a shared, cooperative interest in the success of the coffee shop. As employee and shop owner, they also have competing interests. It's in the interests of the shop owner to pay less, and it's in the interests of the employee to demand more.

Revealing that, based on his working hours, the shop's owner is only making roughly an extra $3.57 per hour (based on the store's profits) instead of something closer to an extra $30-40 per hour does not mean that the shop owner and employee don't have competing interests. However, it does reduce the stakes. The amount of effort that it's reasonable to put in to a fight over $3-4 is a lot lower than the amount of effort it would be worth putting in to fight over $30-40.

The $3.57 difference also suggests something important: the best way to make more money would be to leave working at the coffee shop and move to a higher-paying position in another industry. No amount of labor organizing is likely to convince the coffee shop owner to pay wages high enough that he loses money on the coffee shop every month.

12. Emerging Challenges

The 21st century is a time of increasing capital sophistication and high energy production. New technologies are likely to change underlying dimensions of human existence – and this carries some risks with it.

12.a. Artificial Intelligence

Computers increase the dimensionality of the response in a capital system, making them more labor-like, and less capital-like.

Current artificial intelligence systems have exciting new capabilities compared to systems even from 2014, and are massively more advanced than what was marketed as AI 2004. However, they are still limited. Systems such as resume processors may consistently give wrong answers, and humans may have to reshape themselves or lie to fit the absurdity of the AI system.

Much of the danger, however, lies in the other direction.

In increasing the ability of automated systems to listen to and process information in natural language, AI systems may enable levels of state surveillance that were previously impractical due to a higher required ratio of surveillance workers to surveilled individuals in the past. This previous lack of surveillance capability may have formed a natural defense which allowed the private discussion of dissenting viewpoints, even when it was prohibited by governments or socially suppressed. Many ideologies view censorship and coercion of speech as legitimate, but the limits of these ideologies (their lack of dimensionality compared to the world) may introduce severe mismatches between the world and what the ideology says – enabling them to more thoroughly censor all dissenting speech by massively parallel monitoring of citizens may lead to disastrous outcomes.

12.b. Genetic Engineering

Markets are likely to incentivize genetics that lead to economic success, such as genes that lead to working hard for long hours, while disincentivizing other aspects of being human that are important in other parts of our lives and to society at large. It may become more difficult for people who are similar to present neurodivergent people (as they currently exist) to find employment and support themselves, which would be bad for them individually (and thus morally questionable for this reason), but also bad for the poorly-understood and financially-uncompensated functions they perform for the broader human society.

States have insurance-like reasons to want to reduce impulsiveness (as many crimes are impulsive and not particularly financially sensible), and political leadership have reasons to want to reduce dissent.

Socialists have incentives to try to flatten the distribution of talent, as this would make it easier to flatten the hierarchy of society, and prevent uncontrolled outliers.

Other ideologues will have their own reasons to mess around with humanity's genetic code. Some of them may use justifications to support this that we haven't yet imagined, or which are not currently widely known.

Whether these things can actually be done genetically, the degree to which they can be, and the costs involved, are open questions. However, the incentives are there. Narrowing humanity's genetics down too much will reduce humanity's adaptability, and create exactly the kinds of vulnerabilities that short-termist politicians and their supporters don't like accounting for until they are already a problem.

12.c. Peace & Justice in the Future

The purpose of this essay isn't to offer a complete answer on what to do about these issues. Rather, these technologies, and the policies surrounding them, will be an important thing to keep an eye on as we head into the middle of the 21st century.

The ideologies of the mid to late 21st century are likely to be significantly different from the ideologies of the early 21st century and the ideologies of the 20th century. Many 20th century and early 21st century ideologies treat human genetics as something that cannot be changed, not only in terms of what policies make sense, but also in terms of which topics are acceptable to discuss. There has been significantly more discussion of artificial intelligence and its associated risks, but late 20th century thinkers had little in the way of legitimate artificial intelligence to work with.

There will be a lot of work, much of it political and ideological, coming up in the future. To make and keep peace, while retaining the rich nature of human existence, and maintaining justice, will depend on many choices that will be made by the current generation (and their children).

I hope that the information provided in this essay will help them, and help you, to achieve that.

-

[1] Given the expense of a car, taking 10 hours to decide on a car would actually be reasonable.

An Example Problem

Suppose there is an island in the South Pacific where a new hospital is being constructed. There are two ethnic groups, the Reds and the Blues. They get along fine, but in accordance with computer models that show self-segregation with even weak preferences for similar neighbors, most Reds have moved to the north of the island, and most Blues have moved to the south.

A new hydroelectric power plant will be constructed to power the hospital. There are two potential suitable sites, one on each end of the island. Each site has stunning natural beauty, and quick access to it provides nearby islanders on that end of the island with a benefit. The sites are extremely similar for engineering purposes. The location of the new power plant will be put to a vote.

It is expected that the mostly-Red northerners, who have a slightly higher population (at 51%) than the mostly-Blue southerners, will win and impose the plant on the Blues in the south.

Where should the democratic boundary be placed?

While there is a Utilitarian argument to build the power plant in the south, as the impacted population is slightly smaller, if the democratic boundary is 'the whole island,' then the mostly-Red northerners can continue to vote for their own advantage every time one of these questions comes up, even when their preferences are more weakly-held than those of the southerners.

If the island is divided north-south geographically, each section will have an ethnic majority that could make laws that are poorly-suited to the local ethnic minority, such as the southerners using tax money to pay for Blue churches. Further, there are significant questions as to how the island will be divided geographically. Two polities? A geography-based legislature? Federalism? And the more complicated the solution, the more work will be required.

Splitting the island ethnically, with different laws for Reds and Blues, would require a means for law enforcement to quickly identify (within moments) whether someone is a Red or a Blue - and wouldn't resolve matters where Red and Blues interact but their laws are not in agreement.

Solutions?

Complicated systems of government inter-relating geography, ethnicity, and democracy are not a hypothetical. It's especially not a hypothetical on an island - Britain brought Indians to Fiji, and a system of legislative seats allocated by ethnicity was implemented as a compromise between their descendants and ethnic Fijians.3☆

There is no right answer to our example problem. (Even a relatively libertarian solution will have similar problems with disputes over what counts as an externality and how damaging those externalities are.) While some answers will be overall better than others, that doesn't mean that there can't be multiple answers that are equally good (or equally bad), with no clear way to choose between them. Even the slick technical solution of spending more money to build solar panels and batteries rather than a dam involves having less of some other resource instead.

-☆☆☆-

[1☆] Corporate Finance Institute: "Externality"

An externality is a cost or benefit of an economic activity experienced by an unrelated third party. The external cost or benefit is not reflected in the final cost or benefit of a good or service. Therefore, economists generally view externalities as a serious problem that makes markets inefficient, leading to market failures. The externalities are the main catalysts that lead to the tragedy of the commons.

[2] This is meant in a broad sense of "minority" - in a fraternity in which 51 members consistently vote to party loudly while 49 members consistently vote to study quietly, the quiet members would be a consistent minority even if they aren't racial, ethnic, or sexual minorities.

[3☆] Wikipedia: Fiji # Responsible Government

The Problem of Police

Police departments must always work with incomplete information. This applies both to who they arrest - at least some suspects will later be found innocent and released - and to how they arrest - a man being arrested might have a weapon on him or might start behaving erratically at any moment. Physically capturing people without harming them is difficult due to the forces involved, which is part of why weapons like tasers, pepper spray, and plastic bullets are now described by some as "less than lethal" rather than "non-lethal."

Given a large enough number of encounters, even a department composed of entirely-dispassionate military robots armed with tranquilizer darts are going to seriously harm someone, and most police departments are much worse than this. Police violence is a notorious problem, and reports of police corruption or covering for bad behavior aren't far behind. Communities must choose between the risk of getting harassed and murdered by the police, and the risk of getting harassed and murdered by criminals, with better departments doing better at both metrics, but perhaps at a cost to some other factor we care about.1

Mapping Out the Problem

I'm not a criminologist. I don't know exactly what makes a good police department - but I bet there are a lot of experts who do, who for reasons of incentives aren't being listened to right now. (Though with recent events, perhaps people are paying more attention.) What I am about to propose is a more general technique which would require further testing.

We could try gathering metrics, but metrics can be gamed. How do we even quantify traits like 'friendly' and their associated trade-offs?2 We could try electing the police chief and top-tier officers, but people already seem to be paying more attention to national politics than to local politics; adding even more local elected officials to memorize won't help that, even before we account for whatever bizarre distortions political party affiliation would have on the cops. (In fact, it's long been said that 'tough on crime' stats for elected officials are driving wastefully unmerciful prosecutions and other bad policing.)

If the department culture itself is corrupt or has networks of corruption, we're even worse off. Technically, we could ax the entire department and replace it with a new one - Camden, New Jersey, replaced their city department with a county one - but this is likely to result in some disruption and to require a lot of political energy.

Dividing the Array of Cops

...but what if it were easier to just get rid of an underperforming department? What would that require?

First, to prevent service disruption, we would need a backup department that is ready to work within weeks or months. Second, whatever structural factors corrupted the first department might corrupt the second department, so we'll need a stream of replacements.

For raw numbers, let's use New York City. (I don't have anything personally against New York or the NYPD. New York is just really big.) The city of New York has over 38,000 police officers for its population of about 8.3 million residents (a ratio of about 216:1). It would be difficult to replace an entire force the size of a town in one year and ensure quality, and it would take several years to rebuild institutional competence.

However, the challenge of replacing the NYPD doesn't come from the 38,000 personnel overall; a force that large might need to replace 1,000 personnel a year from retirements alone. What makes us unable to replace the NYPD is that all 38,000 cops are in one department. If we had a larger number of smalller departments, then the number of personnel we'd need to recruit in one year to replace one department would be much more manageable.

Charter Justice

At this point, a more libertarian writer might suggest that if we're replacing the NYPD with smaller departments, then we should replace it with competing private security services companies ('PSCs') and issue security services vouchers. I'm skeptical of this approach for a variety of reasons.

In the voucher scenario, only living friends and relatives of a homicide victim are still customers, as the homicide victim himself cannot be a future source of revenue for the PSC.

Regular police departments are already under political pressure to ignore crimes by the wealthy and politically well-connected. If people can pay more than the voucher amount, the department has further incentive to do so.

Regular people are already tasked to capacity on obtaining, verifying, and processing information, to the point that our society is under epistemic stress. Searching through and verifying the records and reputations of a dozen PSC firms is unpaid work that we can expect will be done poorly.

The ideal coverage zone for PSC firm competition is the individual or the building. Neighborhood-level coverage organizations would create principle-agent problems and start coming up with their own de facto ordinances to save money - but without neighborhood-level coverage organizations, it would be difficult to gain fast responses and economies of scale.

Under a non-profit department model, it is more difficult to siphon off resources that would have gone to policing as cash for personal use and the return on investment for doing so is lower (due to costs of evasion/laundering). A for-profit department is incentivized to drop the least-profitable cases and to minimize expenses while externalizing the costs even more than a conventional department.

It is not that the impulse for private firms is wrong in the general case, but rather it's a matter of the incentives in the specific case. Private firms do fairly well at producing ordinary goods and services with repeat buyers where customers are easily able to assess the quality of the good or service. Policework doesn't fit the characteristics as well, in part because it's about enforcing (and thus influencing) the rules themselves.

For these reasons, I think competing public departments would be a better choice than a private security model.

Departments

How many departments should the city have? When a department is larger, it can gain economies of scale on paperwork, storage, and so on. When a department is smaller, it's easier to replace and there's more variation, but the more departments there are, the more departments there are to keep track of. We must manage the flow of information.

For many cities adopting this multi-department framework, five (5) might be the optimal number of police departments. In New York, each would have around 7,600 officers. (A meta-department could be placed in charge of procuring city-wide police software and standard equipment.) Three departments would be too few - each department would only have to be better than two others, and a department closure would raise the other departments' workloads by 50%. Seven or even nine departments might work for New York (with around a million residents per department), but every jurisdictional boundary that needs to be cooperated over for an investigation of a cross-jurisdiction crime adds more friction and makes the investigation less likely to succeed. Smaller and more temporary task forces could be created between departments for city-wide crimes.

Once we have five departments, it becomes easier for the City to remove one should an entire department be implicated in a Nepalese tiger smuggling ring, or some much more ordinary kind of malfeasance.3

But how often do we actually want or need to remove the entire department at once? We might want to decrease a department's power and reach without totally removing it, for less disruption.

Territories

Instead of dividing the City of New York into 5 territories, let us divide the city into 100. To start, each department gets 20 territories, all contiguous. Funding is on a per-territory basis, based on demographic factors such as population and level of poverty.

Who knows whether a police department is doing a good job? How do we solve the question of what metrics are appropriate and what the right balance of risk is? We don't. We punt that question to voters. Every two years or so, all police territories that are either on the border of a department's controlled territories, or are adjacent to a border territory, vote on whether to switch to neighboring police departments.

The voters don't need to keep track of much. Most years, the main question they need to answer is: "is my neighborhood's police department doing a better job than the police department across the street?" They don't need to know who the police chief is, or what the department's exact policies are, although of course informed voters can do research on the departments to find out.

City elected officials manage the map requirements. Only 5 territories may be transferred away from a department in one round. The area controlled by a police department must be contiguous. A transfer may be refused by city officials if it violates one of these two conditions. The actual process of switching will occur over a period of 6-12 months.

Departments that are well-regarded will gradually grow, while departments that are hated by the citizens will shrink. If they are unwilling or unable to reform, a department will be dissolved when it falls below 5 controlled territories.

For more immediate results, the Mayor can put a referendum to dissolve a department to the voters for the following election.

New Departments

To stay at a total of 5 departments, a new department must be added whenever an old department is dissolved. Where do new departments come from? To create a new department, the largest (and thus presumably most well-regarded) department is split into two departments over a four-year period. Alternatively, a vote may be held to find a critical mass of territories that want a new department, and a new department can be formed there. The decision on which of these two processes to use will be left to city elected officials.

Departments can also be allowed to move beyond municipal boundaries, allowing the movement of departments from the suburbs into the city, or between suburbs.

It's quite possible that the new department procedure will never be used. The threat of losing control of territory, and thus losing jobs for officers, and losing prestige for the department, may be sufficient to get about as much performance out of the police departments as they are realistically able to provide with the budgets they have.

With both a source of new departments and a loss of old and underperforming departments, and the movement of departments along a graph of connected territories, we can call this overall approach GRID - Graph of Redundant Inexpensive Departments4 - until a more suitable name is devised. "Departments" are of course a bureaucratic construct. Officers may end up working for different departments throughout their careers.

The marginal employee typically needs a company more than the company, which is much larger, needs the marginal employee. Unions allow a balance where 'employee' bargaining organization size is in the same class as firm size. Under GRID, there should be one union for each department, to prevent the police unions from overriding city policy, while ensuring they have firm-size negotiating leverage.

Risks

How much discretion in setting internal policy will the departments have? This will be set by the city elected officials, who may allow more or less discretion as they deem appropriate. The degenerate state is similar to having one city police department, but with higher overhead.

How much will the departments campaign, and how much will the policing become politicized? It's likely that the neighboring territories will be relatively similar to each other, so police policies probably won't wildly diverge. As for campaigning, it isn't just important to do a good job, but also to ensure that others know that you did a good job. The best campaign for a department is probably to have good relations with the community.

Will departments still attempt to cover up failures? The organization is smaller and thus less able to endure the consequences of a major failure. However, I expect that most major failures start out as a minor failure that is then compounded. With the tighter feedback between actions and consequences, and with fewer people to monitor, the intent is that minor problems will be found and addressed earlier, before they have a chance to become major problems.

Risks: Will the Gain Be Sufficient?

The GRID approach only makes sense if police departments aren't performing near their maximum potential given their available resources (such as equipment, talent pool, conditions of the areas they're patrolling, etc), and the reasons for this are within the potential control of the department. (For instance, if a city's tax base evaporates and the police department is used to gain revenue for the city government through fines, it will be very difficult for the department to have any kind of good relationship with the community.)

Creating multiple departments means having more department chiefs, more buildings, and periodic moves between buildings. The total number of cops is unlikely to change, but the total cost of the program might rise by 3-10% (to blindly guess). GRID should only be enacted if we expect that there is a double-digit gain that we cannot achieve through smaller, simpler policy interventions, or other alternative classes of policy interventions.

General Applications

The use of police departments is only a specific example of what is intended as a general technique. If institutions are essential to the maintenance of advanced civilization, but there are risks of long-term institutional decline, then it may be worth breaking them up into smaller units so that one institutional failure doesn't wipe out society's full institutional capabilities in that institution's class. (Further, the rate of institutional decline is likely to vary, allowing us to maintain a higher average institutional performance for longer through replacing the worst-performing institutions.) While the failure of small institutions may be more pro-cyclical on the small scale, an early small institution failure may be more recoverable than a late large institution failure, as it may occur while more resources are still available, and re-capitalization cost to create a replacement will be lower.

On the other hand, large institutions can better survive shorter and smaller disruptions. GRID should be conceived as part of a general discourse on institution size as part of society's ecosystem of institutions.

-☆☆☆-

[1] - For instance, many colleges use 'Affirmative Action' policies to alter the demographic distribution of their incoming student body, generally justified as either increasing the diversity students are exposed to, or making up for disparities earlier in the educational pipeline. Increasing the training standards for police to match that of a 4-year degree might improve officer performance, but will likely require an affirmative action program due to whatever disparities are present in the pipeline for a conventional 4-year college degree, and will likely also increase individual debt and expectations of officer pay. Cities may find they can't lower the pay even if they want to, as they compete for the same pool of otherwise college-bound prospects as many industries do. This may further hinder the ability of impoverished cities to hire an adequate number of officers relative to the number of calls, or drain municipal budgets that could have been directed towards prevention or other programs.

[2] - "We must measure some proxy for officer friendliness. Officer smiles per pedestrian-minute is such a proxy, therefore we must measure it," and absurdities of similar magnitude might follow.

[3] - A tiger smuggling ring is less realistic and used as an example because it's less depressing than a more realistic one like ties to organized crime. On the other hand, in this absurd year of 2020, a department might be found guilty of something equally outlandish.