I just want to mention that many commenters here are reading much more into the name politically than do the game theorists/computer scientists who actually do research on it. The research is focused on the game theory and not meant to make any sort of political points or value judgement. The main question asked is how to design systems so that even in a game-theoretic equilibrium, the total "welfare" is close to the best possible.
Ding ding ding. The Price of Anarchy originally described the surprising result that internet routing seems to work despite the lack of a central planner. [1] The "central planner" is a theoretical maximum for the routing problem - it represents an oracle that automatically knows and applies an optimal routing solution. It's not some kind of government.
It's kind of hilarious seeing all of these people commenting about how this is "utopian" thinking, ancap-blah-blah, and "who says the central planner would be just?" People read what they want to read, I suppose.
If you do want to read something interesting about the price of anarchy, check out this other paper [2], written by one of the original authors, which shows that the PoA = 1 (that is, there is no cost) when there are private "owners" of the edges in a routing network who are allowed to charge a "toll" for traversing that edge. Makes you wonder if a globally-standard-but-locally-owned electronic tolling system could replace gas taxes to fund our roads in a fairer way. This could remove the undue toll burden shouldered by certain unlucky people who have to commute across a heavily tolled bridge. (Seattlites should know what I'm talking about).
> It's kind of hilarious seeing all of these people commenting about how this is "utopian" thinking...
I suppose you're referring to me, since I'm the only one who has used that word.
To clarify, I was replying to a comment, not making any comment directly relevant to the paper or PoA.
That said, if someone applied this paper's results directly to a defense of anarchy, I would call that application Utopian because I don't think game theory provides a particularly compelling model for all of society. FWIW such an application certainly wasn't the author's intent, and also is not what was going on in that thread.
> Makes you wonder if a globally-standard-but-locally-owned electronic tolling system could replace gas taxes to fund our roads in a fairer way. This could remove the undue toll burden shouldered by certain unlucky people who have to commute across a heavily tolled bridge. (Seattlites should know what I'm talking about).
This is an interesting idea (which I think is the role that mathematical models applied to social systems should play -- as a stimulant for interesting ideas and possibly a way of bounding the possible. A basis from which you rapidly diverge, which is OK. TBF I think this is what is also going on in the other thread...)
By "standard", do you mean "every mile is taxed equally"? Or "according to construction/upkeep cost"?
The latter seems hard to calculate and easy to fabricate. I think in a realistic system you'd just always end up with really high bridge tolls. In that sense, I think this application of this paper is Utopian in the sense that, when applied to the toll problem, it makes unwarranted assumptions about accurate pricing.
If by standard you just mean "equal", why not just use gas taxes and set them so that tolls aren't necessary? They approximate distance traveled pretty well, so should be pretty identical to tolling each mile equally. With the added benefit of also indirectly taxing pollution.
Nothing personal about your comment particularly - I agree with how you're characterizing it. I was reacting to the instant jump to political thought people were making when the PoA has plenty of interesting implications in the merely technological realm :)
> By "standard", do you mean "every mile is taxed equally"? Or "according to construction/upkeep cost"?
Perhaps confusing wording on my part - by standard I meant that the system that actually detects the car and charges the dollars and cents is standardized. In other words, a frictionless lingua franca for tolling that exerted negligible inconvenience on the driver as they were driving. Think "everybody has a Visa", not "every item costs $1".
The tolls themselves would be priced neither uniformly nor "at cost" to offset construction expenses. Rather, they would simply be set to maximize profit for the local municipality that owned the tolling machines. Pricing the edges to maximize profit minimizes congestion, and theoretically does exactly as well as a super-deity oracle with perfect information that could plan everybody's route from start to finish. (This perhaps counterintuitive result is described in ref 2 from the grandparent). Whatever money is made in aggregate is used for all repairs and maintenance, so the idea is that small back roads would help to subsidize the cost of bridges, which are disproportionately expensive to repair compared to their importance in the road network. More importantly, the price to drive on any road is directly reflective of the externality you impose while driving on it. Right now, the expense and complication of exacting the toll is high enough that we only extract tolls at critical points like bridges, which means we only use tolls to discourage congestion in these few places. We set the tolls too high there to offset the fact that there should be tolls on feeder streets and other areas that are getting backed up and contributing to the overall gridlock.
In practice, you'd have to get around some issues surrounding the advertisement of road prices to the driver. The system would have to strike a balance between being fair and fatiguing drivers with too much shopping choice. In this respect, perhaps you could be right that there is some idealized thinking going on. But it is just an idea, after all.
This reminds me of the sunpass/turnpike system in South Florida. A toll is charged to your account by driving through toll booths located at entrance and exit ramps (done through an RFID chip or something similar appended to the windshield). The tolls generated are used towards maintaining that highway. It tends to be much less congested with more considerate drivers, compared to 95.
What you are saying is true and crucial to understand. The same observation applies to a lot of named concepts in academia (in voting/game theory and economics especially).
However, I also think that results in these fields can be useful in political discussions, as long as they're used as thinking devices rather than arguments in and of themselves.
The map is not the territory, but "price of anarchy" is a suggestive name. The name suggests, purposefully or not, that this set of mathematical results about an abstract game theoretic model might be useful to hold in mind while thinking about political systems. And I don't think that's necessarily a bad suggestion.
Then they should use less political and value judgement entailing language.
That they don't indicates that they enjoy the attention doing so brings to their otherwise dry mathematical endeavors. Or that they aren't very good at rational decision making. In which case it's not clear that they are the best people to be writing about it.
Yeesh. Remind me to never publish a paper that gets mentioned in a Wikipedia article and then posted to Hackernews 15 years later. I wouldn't want somebody to knock on my door and demand that I find a substitute for the term "Curse of Dimensionality", being that there are no actual witches involved.
Having no intention to be political just means they uncritically reproduce aspects of the ideology they passively go along with. Associating anarchy with Bad Things was a deliberate government program to suppress its movements(quite popular at the turn of the previous century). Now people do it reflexively and that is very much political. The political is not merely intent.
I think that this is a really interesting concept, but dislike the name. All social anarchists (anarchist-communists, anarchist-syndicalists, even mutualists etc) would see anarchy as a social process rather than a purely individualised "every man for himself" system. It disregards the possibility for non-coercive collective forms of decision making.
but, saying this, a concept that discovers the loss in a system where each actor is purely self-interested is useful and interesting.
> ...as they would like it to become... Friendly society... DRO...
I think this has always been my primary gripe with ancap -- it's Utopian. And all Utopian thinking has the same problem -- it's either irrational, or very rational but starting from very strong axioms. Ancap writers, to their credit, tend to care a lot about crafting strong, rational arguments.
When I think about ancap (or any political theory that sets off my Utopia alarms), I reason through the entirety of society. I raise a lot of objections to myself, and make a note each time I come to the conclusion that I need some assumption about how people will behave.
Then I take those assumptions, and apply them to various state-based systems. Usually I end up with an equally Utopian world, or only require a small delta in the axioms to end up at a Utopian world. To me, that suggests that it's the axioms that are doing all the heavy lifting. If you can make enough assumptions, the state is much more than unnecessary -- it's also irrelevant! You get a Utopia either way! The enemy isn't the state, it's human behavior.
FWIW the same exercise can be used to derive ancap out of a thorough analysis of communist literature. And it also applies to more centrist policy proposals that posit Utopian-level returns.
At least, that's my experience with struggling through ancap literature. YMMV.
I don't think its all that Utopian. An-Caps take a lot of care to make there institution prove against human weakness. There is also a lot of study into historical examples that explain part of how a An-Cap society would work.
I don't one can expect more from any group. An-Capism is as un-utipian as you can get while still having some vauge definition of an ideal system.
Most AnCaps happily admit that the assumed society would not be near perfect. In fact, I would actually say that it would probably not be that much better then what we have now.
I don't think the axioms are that special, its basic rational choice political science/economics that is applied in most AnCap arguments.
> An-Capism is as un-utipian as you can get while still having some vauge definition of an ideal system... Most AnCaps happily admit that the assumed society would not be near perfect.
Perhaps. That doesn't make it any less Utopian in my book.
> I don't think the axioms are that special, its basic rational choice political science/economics that is applied in most AnCap arguments.
Depends on the writer. There certainly are writers for whom this is true. IMO it's not true of Molyneux, to name one.
Of course, rational choice is not special, but it is an enormous assumption that we're pretty sure is not a realistic description of how people actually behave...
Then I don't understand your definition of Utopian or how it is useful.
Molyneux is no longer an AnCap, he has gone of and is basically something on its own now.
There are actual economics/poetical science PhD working on this stuff, see for example Peter Leeson, Bryan Caplan.
> Of course, rational choice is not special, but it is an enormous assumption that we're pretty sure is not a realistic description of how people actually behave...
Rational choice in this context means not Homo Economicus (as in mathematical maximisation of expected outcome) but rather rational choice limited by information and so on. This assumtion is much weaker and applies to enough people as to make it useful.
Its basically what much micro economics and political science already does. I don't know what better scientific disciplines we have to make better evaluation of theoretical system of humans.
You were likely thinking of moral and ethical lapses, but there's also the crucial problem of IQ and disability. Could a ancap community of PHD students work? Probably. Elsewhere? Well...
Humans as they are now, not some PhD. Again, its that same analysis used when talking about democracy or any other system.
You can go back to David Hume. How can we design democracy so that we can limit what the worst people can do.
This exact same way I approach any system. The reason why I think AnCapism might work, is exactly because I don't think it needs any change in human psychology, IQ or whatever else.
In other words, solve a minimax problem over human behaviour. Good luck with that - we do not even have strong scientific notion of how people make decisions in various conditions on micro level.
AnCap likewise does not work, because we do not know when people would renege on their contracts (essentially steal, injecting accumulation of wealth due to lack of information about what was stolen). Not to mention it does not start from zero and accumulation of wealth is obviously easier when you already have it and becomes easier.
Result is a state = a monopoly, oligopoly or a cartel.
Such a result is only avoidable if somehow economics is not a zero sum game and individual progress/gain is quick enough to outweigh all such initial tendencies, which would only happen in the singularity.
An alternative would be an utopian uncorruptible oracular jury with unlimited power = God.
Hacker News is not the right place for such a discussion.
We do not have perfect information on everything, but we still need to do the best we can to make informed choices. So I use the same we use in standard Science.
There is a ton of literature on law in non-state situation, from legal, political and economic scholars. I suggest you look at it. Its the basis of AnCapism.
Well welcome to nearly all political theory. There is a reason we are still stuck with some fairly well known different systems for all these years.
Human bahaviour is the same so far. And saying that we need to modify human behaviour to make it "better" is a complex road too... The thing is, we tryied to did it with the widespread education of since the 50s, with a more or less loose objective and set of policies, implemented differently in different countries.
We now have more data than ever... But can we really shape our future toward something that help our species to survive?
I think that it is the big hole in the set of goal that Elon Musk set for himself, and it talks also a lot to the goals that sama want for YC. How do we educate, teach, create opportunities, so that humanity keep going or have the tools to overcome what is to come...
> Well welcome to nearly all political theory. There is a reason we are still stuck with some fairly well known different systems for all these years.
There are plenty of anti-utopian and materialist political theories. The real reason why we've seen only a few different systems is that they are the ones that most effectively wield power and violence to maintain their class relations. That's the position of Marxists. Most strains of anarchism are also anti-utopian.
I think it is naive to say that the reason we haven't progressed is because we don't have the right set of ideas. Idealism itself is an obstacle. Understanding the material forces that create and propagate our world is essential. Most mainstream political ideologies pretend the world is driven by ideas when it's the other way around.
> Err, by "very strong" did you mean "makes (a conjunction of) many claims" rather than "resilient"?
Kind of. I certainly did not mean "resilient"!
when speaking of axioms, typically "strong" means "a lot of stuff follows from the axiom". It's in reference to the deductive power of the axiom, usually when taken as part of a larger system of deduction.
So for example, "False" is a really, really strong axiom in most systems of deduction.
Strong axioms aren't necessarily a bad thing. But if you need a lot of strong axioms, you might be saying something about a very specific hypothetical situation...
"Anarchist" is antipodal to the "nationalist" part of "nationalist mercantilism", which is historically associated with "capitalism" in such a way that many people confuse the two.
This is exactly why some people prefer the term "propertarian anarchist" to "anarcho-capitalist", so as to avoid arguing over the meanings of words rather than the tenets of the school of thought.
An anarcho-capitalist cannot make and enforce any law establishing rights in property, but instead recognized that by societal convention, people can claim objects for their exclusive use, and defend such claims by voluntarily respecting the reasonable claims of all other participants in the society.
If you have faithfully watched the television series of The Walking Dead, "The Claimed Gang" that was featured in season 4, episode 11, were basically anarcho-capitalists. If you said "claimed", whatever it was became yours, and if you violated someone else's claim, you got a beatdown from the whole gang.
As far as I can tell it's the only consistent form of anarchy. For instance in discussions with anarchy-communists, if I were to develop a machine that manufactures widgets in my spare time at home, that machine would have to be seized since communism is the public ownership of the means of production. The communists I've talked to believe that there would be some public police force. When asked how any decision would be made by the collective, since it is a system run by "society" not by any individuals, the answer often hear is through democratic means (i.e. not anarchy). Considering only those two aspects, you already moved from anarchy to a form of democratic socialism.
Edit: If you have a different point of view, please tell me where I err. The view point I've developed is based on my discussions with the communists, but it obviously doesn't mean that I've gotten the full story.
Is your magical widget machine shitting bricks out of thin air?
If it needs input, energy, resources, suppliers, other people for developing ideas around it, then its not "your own making", it is already a community project, you can keep it at your home, its produce will have to be spread to other people like they are spreading energy and resources to you, a community of sharing and caring, and not this braindead capitalist ayn-rand bullshit "my own shit is shinier than yours".
The communists are a bunch of their own, why not just ask with anarchists?
> if I were to develop a machine that manufactures widgets in my spare time at home, that machine would have to be seized since communism is the public ownership of the means of product
No. That's personal (not private) property. The MoP are manned by labour supplied by the proletariat. If you are having people working with that widget maker (aside from yourself), and you are paying them wages (or slavery), then it qualifies as MoP and the workers would seize it.
In the same way, a programmer's computer, provided the owner is not using it as MoP for wage-labour, belongs to the programmer.
>When asked how any decision would be made by the collective, since it is a system run by "society" not by any individuals, the answer often hear is through democratic means
Our experiences differ; usually the response is that people can organise themselves freely into collectives or communes. There is no police force to force you to join a commune or obey their rules.
But if people can freely organize themselves into collectives or communes, then some of those collectives will be more successful than others, and without mandatory redistribution, you'll have inequality again. Also, if people can leave collectives, then inevitably there are going to be collectives with few or only one single member left, and you're back to individual ownership.
A lot of communists posit that capitalism (and perhaps in some cases statism) cause and/or arise from something kind of resembling a mental illness. Marx's "Alienation" and especially Delueze's "Desire" have this flavor. I think this is one reason why there's a whole brand of communism that's closely tied to Freudian/Lacanian style psychology.
That psychological aspect usually plays some role in transitionary theories for certain brands of communism, especially anti-statist brands. Although nailing down exactly what role it's supposed to play is usually pretty difficult, it's fair to say that the answer to your question would -- for a lot of anarcho-communists -- boil down to something related to these psychological theories.
This post may sound uncharitable to those communists. And I don't agree with their theories. But I also think that their work contributes something very important to discussion of fringe political theory:
Flipping a psychological switch in the minds of the masses is an absolute pre-requisite for most anarchist theories.
IMO that includes anarcho-capitalism. The communists at least recognized this way earlier than everyone else and are very careful about identifying psychological theories that match well with their sociological theories.
We can argue whether psychological switch-flipping is a realistic expectation -- I'd argue not. But any time I read political theory, I try to identify where the author invokes a form of switch-flipping. 99% of the time the author just makes overly charitable assumptions about human psychology -- that's usually the case in mainstream politics. Communists sometimes have some Freudian psycho-analytical flavored theory. Anarcho-communists like Molyneux prefer weird pseudo-logical reasoning about moral axioms.
Implicit assumptions of switch-flipping show up in mainstream politics as well (IMO "drain the swamp" and "black lives matter" both make implicit assumptions about switch-flipping that under-estimate the stubbornness and importance of human psychology...)
Well, that's a big weakness of these theories, I think. My question is not terribly complex or dependent on unusual or unrealistic situations.
I would also note that sometimes, people operating off a 'psychologist' basis like you've identified, misunderstand the critiques of others as being also 'psychologist'. For example, the Hayekian critique of central economic planning is often characterized as a psychological argument that "people are naturally selfish and greedy". Thus, if you can get them to stop being so selfish, the objection is overcome. However, what the critique is actually about is inherent limits to knowledge and information, which would apply regardless of psychology.
> My question is not terribly complex or dependent on unusual or unrealistic situations.
No, usually the psychological arguments are needed to hand-wave away crucial but obvious problems.
> Well, that's a big weakness of these theories
No, it's not. It's perhaps their most important insight.
I mean, IMO they're wrong, but the insight about the need for a psychological aspect is important and correct.
Libertarians and especially anarcho-capitalists need similar psychological arguments (the fact that they often hand-wave over this need not-withstanding).
Let me clarify: I'm saying that the "well, your objections are just psychological" aspect is the weakness. Sure, to have a psychological theory is fine, but too often it drifts into essentially "objections to our theory are merely based in the same psychological malformation that is produced by the current system - therefore when we get rid of the system, the objections will go away too". This is a weakness, I think.
Except maybe there's a system for which that's really true, and you and I prevent us from ever getting to that system because we reject arguments like this on the basis of form? Oh dear ;-)
> (Most) Anarcho-Capitalism assumes humans as they are now.
No, most anarcho-capitalism that is analytical (rather than holding anarcho-capitalism directly as a first principal) relies on outdated idealistic models of humanity.
> It uses the same analysis as do political scientists and economics use now to model systems.
In that, despite its well known flaws, the rational choice model is still frequently used (either as a tolerable simplification in some contexts, or simply from inertia in others) in both political science and economics, this is true. OTOH, it remains an known-to-be-false idealized model.
Your argument has nothing to do with Anarcho-Capitlaism.
You are simply arguing against most political science and micro economics. I use those disciplines to explain why I think Anarcho-Capitalism makes sense.
If you don't except these then I can not convince you of anything, not even that dictatorship is bad or anything else. There is no point arguing.
If that 'collective of one's legitimately only needs one worker, then that isn't a problem. If someone else joins him later then he has to share the MoP.
But are collectives allowed to exclude people? If not, how will they select appropriate workers? If you can just show up at a collective and say "I want to work here, and you can't exclude me or kick me out" then how will they be able to function? People will just leave collectives with poor returns and make collectives with good returns take them on.
You CANNOT freely organize yourself in an communism.
This is because as soon as you start "freely organzing" a whole bunch of my property now apparently becomes "means of production" and gets siezed from me.
What if I and a group of other people want to organize and work together, in exchange for wages with fully voluntary other people?
>What if I and a group of other people want to organize and work together, in exchange for wages with fully voluntary other people?
There's no problem with that, but the people working for you wouldn't be very clever - there would be no reason to sell your labour under Communism. In fact, it would be so backward, the exploitation so evident, that nobody would even bother. Why would you want to work for wages when there are no wages under Communism?
Nobody is going to stop you from doing it. Communism is stateless. But you'd be pretty stupid to be doing it, and it would be a massive waste of time and resources. Nobody in their right mind would work for wages in a system where everyone else gets the product of their labour.
In this wonderful future, there will still be scarcity. There isn't going to be an infinite number of things.
Sure, maybe all my food and housing is paid for. But what if I want more stuff, that I couldn't normally get?
What if I WANT to work for a wage, in order to get these other things? That doesn't sound stupid at all.
Or perhaps the person who I am working for, due to him having a really good business model, or idea, or skill set, produces so much product of labor that I rather get a small portion of my product of labor from him, than get an even smaller product of labor the "normal" way, even if the normal way I receive is my "fair" share?
I as an employee do not care a single bit about getting my "fair share" of product of labor. I care about MAXIMIZING my absolute amount of product of labor.
As in, I'd rather have 10 percent of 100,000$ than 100% of 1 ,000$. Why would I care if my boss gets 0 dollars or 90,000$s?
I don't care if my boss "exploits" me, as long as it is a better option than working for a "fair" share anywhere else
> if I were to develop a machine that manufactures widgets in my spare time at home, that machine would have to be seized since communism is the public ownership of the means of production.
This is a very wrong view of communism. If the communists you've talked to have told you that this is what communism is, then I'd very strongly advice you to consult some introductory texts on communism.
A common basis for handling this is to discuss what maximises liberty for society as a whole.
If you make a widget manufacturing machine, and it doesn't use much resources, and others can make their own, then there's no reason for society to care that you have your own widget manufacturing machine. Public ownership of widget manufacturing can be done simpy by making another widget manufacturing machine. Liberty is maximized by cloning your machine rather than by seizing it: You get to keep yours; society still gets widgets.
If you make a widget manufacturing machine, and it requires significant use of shared resources - more than your fair personal share - or require more people to operate, then a society that wishes to maximise liberty for all will simply deny you more than your share of shared resources, and deny you access to other workers unless you share control over your widget manufacturing: You are using not just your resources, but the resources of others too, and maximising their liberty involves not handing control over that to you as an individual.
And this is the big gaping contradiction between anarchy and ancaps: Whether you can monopolise resources beyond some reasonable definition of "personal property".
Almost all societies put strict limits on private property because we recognise that enforcement of private property, while it may enhance the liberty of some, will deprive others of liberty. Some societies more than others. E.g. I've in the past brought up the Scandivian countries "freedom to roam" which guarantees extensive access rights to non-built-up private land on the basis that letting a land owner prevent people from walking through their forest, for example, is a massive limitation of liberty on the overall public and only provides very minor additional liberty for property owners in comparison (in Sweden such rights are part of the constitution; in Norway it wasn't legislated until the 50's or 60's because the principle was considered so self-evident it wasn't seen necessary to codify it in law)
So from my point of view, ancap is incredibly logically inconsistent: It wants to enforce property rights, but doesn't want the power structures (the state etc.) are have been emergent from strong property rights.
Okay, I don't like Rothbard, but here I go defending him. The context of that quote is important. It turns out that Rothbard was making the point that arguing over the etymology of these labels is not as important as arguing about actual policies. He concedes that historically those who label themselves "anarchists" have very different and mutually exclusive beliefs than anarcho-capitalists. But he also maintains that they are "without archons" because they oppose coercive rulers.
The entire paragraph of that essay:
> We must conclude that the question "are libertarians anarchists?" simply cannot be answered on etymological grounds. The vagueness of the term itself is such that the libertarian system would be considered anarchist by some people and archist by others. We must therefore turn to history for enlightenment; here we find that none of the proclaimed anarchist groups correspond to the libertarian position, that even the best of them have unrealistic and socialistic elements in their doctrines. Furthermore, we find that all of the current anarchists are irrational collectivists, and therefore at opposite poles from our position. We must therefore conclude that we are not anarchists, and that those who call us anarchists are not on firm etymological ground, and are being completely unhistorical. On the other hand, it is clear that we are not archists either: we do not believe in establishing a tyrannical central authority that will coerce the noninvasive as well as the invasive. Perhaps, then, we could call ourselves by a new name: nonarchist. Then, when, in the jousting of debate, the inevitable challenge "are you an anarchist?" is heard, we can, for perhaps the first and last time, find ourselves in the luxury of the "middle of the road" and say, "Sir, I am neither an anarchist nor an archist, but am squarely down the nonarchic middle of the road."
Anar-caps have always been separated from anarchists. I don't know why they took this name -they were in quite an identity crisis about it [1]-. A better name could be stateless capitalists.
Actually anarcho-capitalism permits anarcho-communist, anarcho-syndicalist, and mutualist enclaves (in addition to any voluntary organization) within of a polycentric legal order. It is the most inclusive variety of anarchism.
Yeah, if someone is willing to enter a contract with you, which they are free to decline.
Anyhow, beyond the mere principle of having to be part of a system you reject, it is quite weird to say that any anarchist could live in an anar-cap utopia, when there are such differences between them ; just like it would be weird to say that any anarchy can fully develop as a community within a statist society.
I don't think ancaps are all like that.
You can't run business if you don't serve your customers well. Does that translate into pathological selfishness and other extreme behavior that goes in customers' nerves? I don't think so, because such business wouldn't last long.
I agree with the GP post above, and I'd also like to point out the idiocy of the "scientific" assumption from Wikipedia that "central planners" have superior information compared to self-interested agents who follow their own optimization strategies. F. A. Hayek is one of best authors on this topic.
>You can't run business if you don't serve your customers well.
Serving your customers well is entirely out of self interest, especially in large corporations; in fact, this is the easiest example of "every man for himself". You serve your customers well because ultimately it brings you more profit. Sure, it's not as sure sighted as screwing over every customer, but people who are selfish aren't necessarily stupid. I would argue that the very act of running a business in such a society is not only wrong due to the exploitation of workers, but also the motivation underlying (almost) every transaction in which your goal is to sell things in such a way as to get yourself the largest profit.
Of course that doesn't happen in real life. People have moral standards. I think I'm just a cynical ancom with regard to how people would really behave in the absence of all regulation. Let's not forget what large companies can still get away with today, and imagine what it would be like with anti-competitive monopolies, oligarchies, and two or three organisations controlling mainstream information sources.
So sure, your business doesn't last long because you were too obviously selfish. Your next plan of action is to pay a news outlet (or better yet, already be in control of one) to cover you. Or if you can't do that, there's still hope - you can band together with like minded people and start another business. Not without worker exploitation, of course. You need all the surplus value you can get.
I simply do not have the hope that "serving your customers well" is enough to prevent what actors in a capitalist system are truly capable of doing.
Let's not forget the sheer paradoxical nature of "anarcho-capitalism" (the inherent class system set up of the bourgeoisie and proletariat is the exact opposite of non-hierarchical relations).
I'm not really a proponent of central planning myself, so I agree with you on that.
> the motivation underlying (almost) every transaction in which your goal is to sell things in such a way as to get yourself the largest profit
And the goal of the buyer is to buy things in such a way as to get himself the largest surplus, i.e., the most value for a given cost. So by your reasoning, all buyers are selfish just as much as all sellers are selfish.
> People have moral standards. I think I'm just a cynical ancom with regard to how people would really behave in the absence of all regulation.
Moral standards existed long before any regulation. We have moral standards because we evolved that way--i.e., because they are adaptive for a species that has to form cooperative relationships in order to survive. Cooperative relationships include economic relationships--specialization and trade. That's how we build wealth.
It is true that, as soon as people start building wealth, there is an incentive to plunder it instead of building more. One way of describing a key problem with many (if not most) modern societies is that they are set up to reinforce, or at least not discourage, the incentive to plunder.
> Let's not forget what large companies can still get away with today, and imagine what it would be like with anti-competitive monopolies, oligarchies, and two or three organisations controlling mainstream information sources.
Um, you're describing what things are like today. And the reason they're like that is, in large part, because one of the things large companies can get away with is buying political power and influence, which they can then use to plunder instead of building wealth. But the reason they can do that at all is that political power and influence can be bought, because it's centralized.
>So by your reasoning, all buyers are selfish just as much as all sellers are selfish.
I don't disagree with this. But I think that there is more harm that comes from the corporation to the consumer than there is from the consumer to the corporation, or at least more potential harm. And I think that arises out of selfishness that is unfortunately built into the system.
>Moral standards existed long before any regulation.
I'm sorry; I wasn't trying to make the point that moral standards have anything to do with regulation, but rather that regulation creates a "bare minimum" for the kind of behaviour expected of the participants, regardless of moral views that very much differ from person to person (and of course from sociey to society).
I think that a relationship can be cooperative, but with one side being favoured much more than the other - the capitalist who exploits a workforce. Of couse both need to cooperate - but it does not mean that such cooperation is fair. It is often accepted because there aren't fairer alternatives.
I wish for a different kind of cooperative relationship, one which I view as more equal and participatory. At the moment, I view it sort of like an EULA.
"Fair" is subjective. Our moral intuitions give us fairly consistent answers for simple cases, but most cases are not simple.
One response to this problem is to point out that a free market is the best mechanism we know of for maximizing "fairness" in the sense of bargaining power. But free markets are actually pretty rare. For example, large corporations' wage structures, which dictate the terms of many people's employment contracts, are not the products of a free market; they are the products of the corporations' internal processes, which are dictated by top-down centralized control. (To some extent they are also products of negotiations, for example with labor unions, but that just extends the top-down centralized control to the unions.) So an obvious way to make cooperation fairer is to decrease the average size of corporations, in order to expose more transactions to free markets. In the absence of regulation, I suspect that this is what would actually happen, because most large corporations are the products of regulation, not of free market competition.
"Fair" is subjective. But from a utilitarian perspective, I think still it would be fairer for the working class to own the means of production.
From a Marxist perspective, a "free market" in which people sell their labour in order to survive isn't fair at all. Whether or not you have a corporation with a big internal centralised form of deciding wages, you still have the product of labour sold for more value than it was bought at.
It might be fairer if corporations became smaller and there was more bargaining power because there is more accessibility for decision in the free market. But I don't think it's fair in other aspects - the aforementioned exploitation (occurring even in the absence of regulation), there is nobody to defend property or even establish the validity of the concept of property, you probably have to pay to be protected by a police force, and all the egregious institutions of today would probably continue, including needlessly expensive health care, there is still the problem of sweatshop labour (which most people, even though informed, don't care enough about to do anything about; imagine how it is if Apple/Nike/whoever were to own a few news agencies too). On top of that, the fact that few exchanges are truly voluntary for many who are less fortunate.
However I can see that there may be more workplace democracy, as a result of internal practices being opened to the free market.
> from a utilitarian perspective, I think still it would be fairer for the working class to own the means of production.
However attractive this might seem theoretically (it doesn't to me, but I understand it does to many people), we have run this experiment in practice and it doesn't work, at least not in the obvious sense of "workers own the means of production". The problem is that "own the means of production" doesn't help unless ownership means control; and in practice, if you have large industrial factories organized with centralized top-down control, it's impossible for all the workers to "own" it in any useful sense. So it just ends up being another vehicle for centralized power.
OTOH, if "own the means of production" really means that each worker owns and controls all the tools he needs to produce, and simply trades what he produces for what other workers produce, then what you have is a free market. In other words, for "workers own the means of production" to be true in any useful sense, every worker needs to be an entrepreneur, basically owning himself and his skills and tools as a small business. I would love to see this happen, but unfortunately I don't think it's likely to on any large scale any time soon. (In some fields, though--programming as a free-lance craft is an example--it can already be true for a significant number of people.)
> you still have the product of labour sold for more value than it was bought at
I don't understand what you mean by this. The value of any product of labor is not determined by the laborer; it's determined by whoever is going to use that product. That's not because of an evil plot by large corporations; it's an unavoidable fact of life. Anyone who cannot produce everything they need by themselves has to trade with others; and that means being able to produce something that someone else will trade for at a price that is enough to compensate you for the labor involved, plus whatever surplus you need to obtain other things you need. There is no reason for the someone else to pay you more just because you used more labor, if the value to them is the same either way.
Everybody understands this when they are the user; if you hire someone to paint your house, you're not going to pay them more if they use a toothbrush to do it. (IIRC pg used this example in one of his essays.)
> there is nobody to defend property or even establish the validity of the concept of property
I'm not sure I understand what you're saying here. The need to defend property and property rights is inherent in the idea of trade; if we all need to trade with others, then we all need to have property and property rights, and have them defended if they are threatened. That will be true whether the average corporation is large or small, or even if corporations did not exist at all.
> However attractive this might seem theoretically (it doesn't to me, but I understand it does to many people), we have run this experiment in practice and it doesn't work, at least not in the obvious sense of "workers own the means of production".
The main large scale experiment I am aware of that meets the "obvious sense" of workers owning the means of production is the Mondragon system and similar labor cooperative ventures, which seem at least modestly successful, even when operating in a legal and political environment in which that structure is not the norm around which most rules are optimized.
There are certainly failed experiments where precapitalist states have been overthrown by regimes in which the state, run by a vanguard party acting nominally in the name of the workers, collectively, owned the means of production, which have failed spectacularly for reasons which may be related to the ownership structure (though other explanations are available), but those don't seem to be operating in anything like the "obvious sense" of workers owning the means of production.
> The main large scale experiment I am aware of that meets the "obvious sense" of workers owning the means of production is the Mondragon system and similar labor cooperative ventures
Um, what about the Soviet Union?
(I'm not familiar with the Mondragon system but I'll look it up.)
> those don't seem to be operating in anything like the "obvious sense" of workers owning the means of production.
If you are including the Soviet Union in this category, you are rejecting the very Marxist terminology that you appeared to be using, since the whole point of the Soviet Union was that the workers would own the means of production. Yes, I know it failed spectacularly; that was my point.
> If you are including the Soviet Union in this category, you are rejecting the very Marxist terminology that you appeared to be using
I think if you read the thread, you'll see that you are using the word "you" quite sloppily.
> since the whole point of the Soviet Union was that the workers would own the means of production.
The whole point of Leninist vanguardism, of which the USSR is obviously the first concrete manifestation, was to adapt Marxist rhetoric to be used in societies in which the perquisites Marx identified for socialism as a step on the route to communism were not met, to justify mechanism which were not what Marx and Engels pres cribed. Leninism is, itself, a substantive rejection of Marxism while adopting it's superficial structure.
That said, Marx is hardly the only thinker (even of his time) to argue for workers owning the means of production, and the indirect manner of such ownership in Marxist socialism -- and even moreso it's Leninist vanguardism adaptation -- is pretty far from the most obvious sense of workers owning the means of production.
The actual workers of individual firms controlling them directly rather than the state doing so in their name is, clearly, much closer to the obvious sense of that phrase.
> I think if you read the thread, you'll see that you are using the word "you" quite sloppily.
Ah, sorry, I missed the change in who I was responding to. I did not mean to attribute to you (dragonwriter) things that someone else said. My apologies for the mixup.
I don't mean to say that everyone acts entirely for themselves, with no action motivated by kindness. I mean to say that in a sytem with multiple people trying to do the same thing you are doing (selling for a profit), you'll have to compete with people with various moral standards and ways of getting their profit. This kind of system, I believe, eventually forces out those who are not "aggressive" enough with their methods, and if not eventually, it will happen over time due to the fact that capitalism requires ever higher and higher returns.
While admirable in itself to act kindly in a capitalist system, it is ultimately detrimental to the success (and thus survival) of the company. There are of course examples of ethical practices (free range eggs for example), but they only continue once they become sufficiently practical.
Capitalism does not require ever higher returns. It requires ever higher efficiency in extracting surplus. Barring monopolies etc., this should be expected to continue dropping as competition gets more fierce.
Other than that, your argument above is basically Marx thesis for the cause of the eventual collapse of capitalism: He praised capitalism for providing that efficiency necessary to provide sufficient wealth to eventually be able to erase poverty, but at the same point argued that this efficiency ultimately means that to stay competitive, salary costs will eventually need to be systmatically pushed down (whether by actually lowering salaries or through e.g. automation) until capitalism keeps hitting crises where the production capacity outstrip demand as the competitive forces throw more and more people into unemployment, eventually pushing people to revolution.
This is also the basis for why Marx unlike many others of the time was cautious in his criticism of capitalists for being capitalists: He points out that the same mechanism will keep throwing capitalists into the working class through failure, and hence capitalists are according to Marx just as much unable to change the system from within as workers.
I guess all those colonial wars and government interventions to build and prop up industries never resulted in higher returns. Capitalism is most certainly not just private ownership of the means of production. That is too reductionist. That is like saying biology is chemistry. Technically true, but misses the point.
Because the means of production are held privately there is a class distinction between owners and workers. Owners derive their profit from the labor of their workers. Workers want to work less for higher wages, owners want the opposite. This becomes the locus of class struggle. Because all society now responds to the needs of capital, the superstructure and the state come to serve primarily bourgeois(owners) class interests.
Considering how many wars and coups have been fought to forcibly introduce capitalism into other countries and forcibly extract their resources and labor, to say that private ownership results in higher returns is almost offensive in how it elides the emergent dynamics of capitalism.
Capitalism and nationalist mercantilism are historically related, but one does not necessarily follow from the other.
Anarcho-capitalists usually define "capitalism" as "societal convention allows an individual to own property for that person's sole benefit". In that school of thought, your ability to own things is limited by your ability to defend what you have claimed. You can't just take your money, use it to buy state power, and then use that to get more money.
The an-cap factory owner does not fear that the workers will seize ownership of the factory. They fear that and that one of the employees will quit and become a leaner, faster competitor. A state-capitalist owner can generally prevent that with zoning and regulation, but the an-cap owner has to create an entirely different worker dynamic.
I am talking about capitalism in material terms as it happens because that's what the post I replied to was doing, not an ideal, hypothetical form of capitalism. likewise if we were having a discussion about the real life consequences about Marxism-Leninsim (like a productive discussion on that topic can actually happen on HN) I would be criticizing the USSR as it actually developed and not, say, as it was described by Stalin's theoretical ramblings.
An-capitalism has never happened and probably never will because it presumes there is an ideal form of capitalism that can outcompete the actual capitalism that exists which is inextricably tied with state power. In every development of capitalism in every country starting with the very first ones in Europe, the state's power was used to disrupt the old order and establish a new one. It was not the only necessary ingredient, but it was essential.
This is exactly why no one who wishes to have a serious discussion should ever say the word "capitalism"--never, not ever--because you will either waste time up front in defining terms, or later on, when different people are using different meanings for the same words.
"Anarcho-capitalism" will never be realized, because it combines two such words into one brand that virtually guarantees that no two people will be able to describe it in the same way.
> You have this awesome outcome, where purely selfish people are still doing the "right" thing.
I don't think it's awesome. Sure, it may not be bad just because of the fact that it is selfish, but I do not think that capitalism produces "awesome", at least compared to theorised alternatives, solutions. In contrast to feudalism, I would agree.
I think that motivation not only sets the tone for the current transaction, but it also influences the sort of relationship that builds over time. While not immediately apparent, it has a great effect on the power dynamic of the relationship. This is readily visible in human to human relationships, and I thnk it can be carried over to the relationship between bosses and employees, the bourgeois and proletariat, etc.
>Serving your customers well is entirely out of self interest
This general tenet starts to break down the fewer competitors operate in the market. Monopolization allows corps to capture supernormal profit and externalize as much cost as possible. When there's only one person that's got what you need, they can treat you how they want.
A common argument against AnCapism is the idea that monopoly will form and eventually there is a sort of superstate.
This is of course a old marxist argument, capital is inherently centralising.
There are however good reason to believe this is not true. In fact, most economist don't follow this line of thinking anymore.
No AnCap would deny that there would be some larger then healthy cooperations but its an acceptable problem that is pretty hard to solve for any system and is often made worse by state based systems.
Well, Comcast's oligopoly is largely established by the fact that rights-of-way and pole-attachment rights are monopolized by municipalities, which tend to approve only a single, or few providers. It's not a natural monopoly.
> Comcast's oligopoly is largely established by the fact that rights-of-way and pole-attachment rights are monopolized by municipalities, which tend to approve only a single, or few providers.
Only because, when municipalities do try to treat bandwidth as a public utility, which they can sell to all providers equally, Comcast and other large ISPs sue them.
> It's not a natural monopoly.
I agree, but I don't think it is primarily local municipalities that are propping it up.
I think it's the structure of rights-of-way, pole-attachments and so forth as monopolies that props it up. The fact that municipalities get sued for trying to sell bandwidth is a problem, but it emerges, I think, because of the fact it's centralized in a government institution.
To some extent, provisioning for bandwidth is a natural monopoly, for the same reason that provisioning for electrical service, water, sewer, and other utilities are natural monopolies. It doesn't make sense to run multiple different electrical, water, and sewer services to the same neighborhood just so multiple different providers can compete; the costs of all that extra infrastructure are too high. The costs aren't as high for bandwidth, but they aren't zero either.
As the size of the "municipality" goes up, though, I agree with you that centralization creates monopolies that are not natural.
>I don't think ancaps are all like that. You can't run business if you don't serve your customers well. Does that translate into pathological selfishness and other extreme behavior that goes in customers' nerves? I don't think so, because such business wouldn't last long.
It can last long enough to make you rich. Or it can last for several lifetimes as long as you have the clout to make it the only deal around (or historical and regional accidents made it so).
In fact, there's also the old saying "there's a sucker born every minute": you just get new customers.
> the possibility for non-coercive collective forms of decision making.
Can you define "non-coercive"? If you want to disallow coercion you must study how coercion actually works on humans rather than decide a priori what constitutes coercion. It is therefore possible that a
"non-coercive collective forms of decision making" does not exist (or can only exist under certain unrealistic parameters).
Alternatively, one could simply pay attention to how life is lived. Most things humans do are done out of habit. One might suspect that all our beneficial habits would break down if there weren't an almighty parental figure lying in wait to punish us, but that also contradicts lived experience: humans in more permissive societies habitually live in relative harmony even more than those in less permissive ones.
Any socialist etc. who studies history will point out that our permissive peaceful society exists with a lot of implied violence that is just made invisible or normalized.
American incarceration rates and health care tied to employer come to mind.
Not sure that "too many people in prisons for victimless crimes" and "tax law that makes employer-tied healthcare almost inevitable" are aspects of society's permissiveness.
His point is that the society might be apparently permissive, at least to those who live in it, but there are parts that aren't (such as those two examples), and people don't generally recognize them as such.
But the comment he was responding to said "humans in more permissive societies habitually live in relative harmony even more than those in less permissive ones."
So the fact that, upon trivial examination, a society is revealed to not actually be that permissive, is not evidence against the assertion that permissive societies are harmonious. I mean, I don't think RodericDay was saying that those two examples were impermissiveness that actually increases harmony.
> One might suspect that all our beneficial habits would break down if there weren't an almighty parental figure lying in wait to punish us
I interpreted this in the vein of a libertarian "we don't need a nanny state". And so I assumed that the person was thinking of the USA as a "permissive society in harmony" and of something like North Korea as an "less permissive society in disarray", as opposed to Sweden as a "less permissive society in harmony".
And so I wanted to challenge the notion that the USA is "permissive".
I'm at least as confused by your comment as you were by mine. b^) On a spectrum of permissiveness, I would have thought we could safely place USA between DPRK and Sweden? I'm happy to consider arguments against such an arrangement... Of course, the idea of a spectrum is also a bit flawed, since there's more than one axis to this.
But we were talking about coercion and authority: GP seemed unable to imagine taking decisions without them. I wanted to point out that most of us are already doing that, to some extent. I hope that in the rest of my life I'll see some progress: more decisions based on personal preferences rather than authority and coercion. To that end, I'm an anarchist. That is, I oppose arbitrary authority.
You're right that it is tricky. Many social anarchists and communists use the phrase 'voluntary cooperation' to describe the social order in a communist society (in the sense of a society without either capitalism or the state).
Whether this is possible or not is beside the point. The argument I was making was simply that the use of the word anarchy in the concept is at odds with its use by most adherents to anarchism
Wow, people are really running away with the political analogy used here.
Calm down folks, they're not ragging on decentralized systems. If you read and comprehend the definitions in the article, you'll see that it's a fully general solution that allows for emergent organization through market systems or cooperation. You can bake those things into the strategies and utility functions of the actors.
It's just a fun and descriptive name for the difference between an idealized central controller (which can force everyone into a configuration that maximizes an arbitrary global utility function) and the lack of such a controller. The outcomes could be the same, they could be different. If you factor in the negative utility of having the central authority in the first place, you can game-theoretically justify your particular favored level of anti-authoritarianism.
If anything, this sounds anti-capitalistic - capitalism is based on an assumption that selfish behaviour of individuals produces greater good for the society.
"It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own interest" - Wealth of Nations, Adam Smith
I tend to think of self-interest being equivalent to selfishness. Your point makes me wonder if self-interest is not equivalent to selfishness, at least to Economists.
Well, what you quote is true, but I think a lot of discussion stems from focusing on the wrong question, namely "does selfishness lead to better outcomes" as opposed to when it does, and when it does not.
Too many people seem to approach economy as an ideology rather than a complex system in which a given approach produces good results here and bad results there, and everything should be subject to optimization and careful design.
> Too many people seem to approach economy as an ideology rather than a complex system
Few of these people have any decision-making authority. This mass ignorance of economics does occasionally bubble up on the policy side. But it is generally, at least in the United States, contained by more knowledgeable (and influential) parties.
Nobody thinks that selfish leads to "better" outcome.
They believe that if everyone acts selfishly, the outcome is still pretty darn good (but of course could be improved if nobody committed crimes or something)
Again, saying that "the outcome is pretty darn good, therefore yay capitalism" is missing the crucial question - when do the good parts happen, when do the bad parts happen, and how to preserve the former while getting rid of the latter.
There is a bit of ideology here, in that there's no particular reason that this tiny little observation about network routing had to be called "The Price of Anarchy". Just the same, it's a mistake to think this topic itself has anything to do with politics or economics. A mistake made intentionally, perhaps, by those with particular political motives.
Network routing is an example. Just because Einstein's original thought experiment involved a train doesn't make special relativity railroad engineering.
Math is important. It tells us the shape of the possible. Game theory tells us when we're creating an unstable set of incentives. Our intuition for fairness is only a crude approximation of game theory.
Societies and their moral systems (there's little difference) don't come out of nowhere.
Sentiments like "game theory is just capitalistic ideology" make me sad. Math, like all sciences, is true whether you accept it or not.
it may be 'true' within the set of axioms on which it rests, but 'wrong' in that those axioms (rational self-seeking behaviour, etc) don't hold in reality. More, those axioms themselves could be reflections of capitalist ideology.
That does not mean it's infallible. Basically, Nassim Taleb stuff. We're wrong all the f*ing time and the more confident and blinded by our models we are, the greater are the effects of our screw ups, there IS such thing as hubris, we've known it long before but it's too abstract to constantly take it into consideration, so we just stop minding. It has happened countless times and there's little evidence we've been able to tame our cognitive blind spots at all, we're just humans, after all.
Isn't libertarian thinking (put very heavily on the edge) the exact opposite?
That a system will become more efficient if actors are let to do things for their own sake, and that actors are smart enough to calculate the negative impact of short term selfish action (that is, actors can help out now, with an exception of a positive roi)
This assumes a theoretical optimal central planner, a superintelligent BDFL of sorts.
Libertarian thinking is that in practice humans are so bad at management that every time we tried central planning it came up even worse than non-optimal Nash Equilibria (which are the failure mode of non-centrally-managed agent systems.) Libertarians would say that no dictator or government in history turned out to know better than The Market. That The Cost of Regulation is higher than The Cost of Anarchy.
Of course, both costs are high for some problems and low for others, and the solution to practical problems usually lies at neither this extreme nor that one.
Some Libertarians are non-consequentialists. They argue that even if the outcome is less efficient, there are moral reasons to oppose central planning. What is interesting is that many argue both points: that markets are more efficient and morally superior. A good example of the distinction is David D. Friedman (consequentialist) vs Ayn Rand (non-consequentialist).
Libertarian thinking actually permits multiple people to agree to not take the most immediately greedy action and to instead voluntarily agree to a mutually-beneficial optimal solution.
I have seen clusters of posts and have not had the time to study: Are they a dialog between unrelated people who share an interest in a topic, or is one person pursuing a theme?
Incomplete models are incomplete. It completely ignores the ways on which the mentioned transport should take place. It ignores that you usually don't have one place from which everybody starts but multiple places to other multiple places.
And if you think about it, individual car mobility is this in practice. The ways are build by someone else but you can decide how and where to drive. And it kinda failed in many areas.
It's not clear what you are referring to. The section on Selfish Routing begins with Braess's Paradox but swiftly moves on to a Generalized Routing Problem which appears to address your concerns.
I'm guessing it's because you're projecting political baggage onto something that is more of just an interesting theoretical result. The Price of Anarchy's original setting was the issue of internet routing, and the "mystery" of how it could work without any central oversight. The answer is that there is a price you pay for having completely decentralized, selfish decision making, but it's bounded. The central oversight bears little relation to a government - it's an oracle that can calculate a theoretical optimum, not some kind of decision-making bureaucracy or something.
You assume the central planner to be the same as a single unit (as in, of the same kind, the same species).
Who is to say that the central planner must be the ontological equal of a single unit? Why not a more complex system?
You could have for example a quorum of single-unit who will need to reach a consensus concerning their diverging self-beneficial impetus. You can also imagine, if you represent the interest of each single unit as a vector in some kind of space, that the quorum must navigate a product of those vectors.
You can imagine an AI then, a machine exploring that vector space that could learn to balance the wants and needs of its single units.
Et cætera. I don't know exactly why you comment was not well received, I am not able myself to downvote comments here and I did not do that, but my own opinion about yours was that it was basing itself on some assumptions that were biased by some real-world considerations that were irrelevant for the concept presented.
"Who is to say that the central planner must be the ontological equal of a single unit"
My impression is that's what most people here are discussing.
I can see a definition of an anarchy in your quorum example, can't you? But you say we could have AI which will find a balance. Interesting concept. Does it imply obeying AI <=> obeying the law?
Well, the quorum could be organized in any matter of way. Expected a consensus for a decision would be akin to anarchy indeed, however there could always be some kind of recursive hierarchy, with another sub-committee of higher authority that would make other units diverge. (Parties within an assembly, with deputies having to follow their party lines). This quorum could also be tied to following the interest of the masses outside or lose substantial advantages, thus tying their own selfishness to the common good.
It could be anarchy. It could be anything else.
The AI could simply do whatever its specialization is. It could be trying to direct the single-units or to disorganize them and try to keep some level of entropy. Maybe its own benefit would be a generalized state of chaos. Whatever.
The system itself could be considered the legal framework. The relationship between any supervisor and its units could be the metaphor for any subservient relationship in the real world, going from managed teams to parenting to politics to social dynamics between friends. It does not mean anything as long as the specifics of the system itself are not defined to permit the dynamics to be correctly mapped on some real-world dynamics.
Obeying the law? Are you obeying the law of gravity? Is it from your own will that you agree to stay on the ground?
No, sorry it was somewhat tongue in cheek. It's just that within the simulation, the way a single unit obeys a central planner (not at all, partially, entirely) is defined by the rules of the simulation, so the rules of the world.
So asking if that rule, that authority of the AI over the units is similar to human law, is I think a bit beside the point. It's more akin to the law of physics. It's simply the fabric of that reality.
Of course, you could then create a simulation that would try to recreate the same kind of domination as within a human society, but that would be only one kind of simulation, one kind of system, with specific parameters. It would not give much about the very essence of the system itself, only of it within constraints.
