= 27d567bc2d48d9f40e9ccc20ea370b15)
ygert comments on Harry Potter and the Methods of Rationality discussion thread, part 18, chapter 87 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (592)
Well, it's worth keeping in mind that the Unbreakable Vow was probably not discovered until well after government had been invented; it wasn't around when people were hunter-gatherers. But I think that saying that people "fell into it" skips the interesting question of how it happened. Suffice to say I'm not convinced that it's as strong an attractor as you're making it out to be. You could as easily say we "fell into" a system where everybody in the world is bound by unbreakable vows to serve a single dictator; why do you think the former system is more likely than the latter?
Also, how specifically do you think that a system running by unbreakable vows would outcompete ones that don't? "Avoid coordination problems" is a black box here. In the starting scenario, we have a society bound by sets of Unbreakable Vows between masters, where slaves (who necessarily have some degree of magical powers, otherwise they couldn't act as binders by giving a fraction of them up) are imperiused into binding the vows, among other societies with different forms of government (probably monarchistic, oligarchic, etc.) The proposed endpoint is that eventually, all societies use this system of governments, because it outcompetes them by being better at solving coordination problems. By what process do you expect this to happen? What coordination problems do you expect them to solve, this way, and how (if you posit that they outcompete other countries in war, for example, what specifically do the people in this society do that makes them more effective in war than what other societies are doing, and why do they do it?)
It sounds to me like the devil of such a system is in the details. "Perfect cooperation, solve all coordination problems" sounds good, but if I think in terms of actual people behaving in characteristic ways, I don't see them taking steps to a particularly effective system by default.
Yes, you make good points. You are probably correct in saying that the scenario could go any number of ways. But my most essential point is this: The Unbreakable Vow is literally a perfect coordination system. That is to say, there is no way of Defecting when you made an Unbreakable Vow to Cooperate. Looking at governments around us, compared to a system like that, they look like nothing more then a hodgepodge of inefficient hacks. People hacked up that (comparative) kludge of a system, because there is nothing better. If there was, I think it's unlikely that they would ignore it and set up a kludgy government like ours instead. In fact, someone from that world would probably find our notions of "government" insane and laughable. (And that person would be right. Compared to a network of Unbreakable Vows, it is.)
My point is, this system is not only better then ours, but so much better that ours pales in comparison. Yes, the devil is in the details, but still,what you are proposing is that that everyone ignores the much better option and settles for the kludgey system that we have. May I suggest that perhaps that is nothing more than status quo bias?
And yes, if the Unbreakable Vow was invented after governments were, the government would try to crack down on this system, to preserve its own power if nothing else. But throughout history, many governments have cracked down on many good ideas, and often the good ideas eventually are accepted. I think that it would be especially true for this system, which is, as I said earlier, simply better then any other possible way of solving coordination problems.
(Although, please note that the system we are discussing (with imperiused binders) is actually a bit more complex, and so a bit worse than a form of the Unbreakable Vow which had no cost to the binder or did not require a binder. But my points still stand, as even though this system is not quite perfect, it still is heads and tails better than the systems of governments.)
One last thing: You asked for details of how society with this network would be better than society with a normal government. It is better in a lot of ways, and here I can not give a full listing of them, but think back for a moment in HPMOR to the chapters entitled "Coordination Problems". I think those chapters give a good feel of just how hard coordination problems are to deal with. For some examples of cooperating problems that would affect a nation, here is a short and incomplete list. (By the way, please note that we have for the most part solved these problems pretty decently with laws and governments. But look at the huge room for improvement. Remember that under a system like this one, these all would be absolutely and completely solved problems. There would be nothing left to discus about them.)
And I am sure that given this list to start with you can think of many more.
But on the other hand, you also can't break a vow that turns out after the fact to be a bad idea. Without it, you can adapt to circumstances and then justify your actions as having been appropriate at the time. With it, if you've made a Vow that doesn't adapt well, you're in trouble. A system of Unbreakable Vows is only a perfect coordination system if the vows themselves are perfectly thought out, which people do not achieve by default.
My issue is that you seem to be assuming that people can just "fall into" a perfect system without giving any details for how they reach that optimum rather than getting stalled at some messy hacks where they are likely to remain due to status quo bias.
What I asked for is not simply a list of ways that such a society, if perfected, would have advantages over our own. I can think of those perfectly well myself. What I asked for was an explanation of the specific steps by which you expect a society would achieve those implementations. For instance, you start out with a regular society, with crime. The Unbreakable Vow is discovered. What steps, specifically, do you believe will occur which result in the endpoint of a society where everyone is bound by a network of Unbreakable Vows to commit no crime?
Just because it seems obvious to you that the end result would be better doesn't mean that people would implement it. In a Tragedy of the Commons, there are often ways that the agents involved could arrange to cooperate among each other (each one could provide collateral which will be returned after a given period if they cooperate, but confiscated if they defect, for instance,) so that everyone will have a greater expected utility if they cooperate than if they do not cooperate, and a higher expected utility than if they did not implement a cooperation scheme. But in practice, people don't usually implement such schemes when left to their own devices. To show that people would be likely to adopt such a system, it's not sufficient to demonstrate its advantages.
Words of honor for parole (think prisoners of war etc.) have historically often served as punishments or forms of security, with the major advantage of being light-weight and minimizing costs and suffering. A huge chunk of all crime is committed by repeat offenders. Hence, crime could be cut by something like an order of magnitude just by making UVs a prerequisite for parole.
This requires no special societal shifts, and is in line with existing jurisprudence using things like ankle monitors to deter breaking terms of parole or committing additional crimes.
From this visible success which will save billions of dollars and millions of lives in the long run, can come acceptance and a slippery slope down to more widespread use - perhaps beginning with application upon simple arrest, much like criminal records can begin compilation these days without anyone whining about state security agencies tracking them (as our ancestors surely would have been angry about, but these days no one can even think of why anyone would object to such state tracking).
I can buy the initiative progressing as far as application of vows upon arrest (although at this point I'm not sure if Ygert is still talking about a system of MoR vows which require the sacrifice of some of a bonder's power, in which case I suspect it wouldn't get that far.) But I find it doubtful that it would progress to the point of everyone being bonded to commit no crimes.
The existence of real life government initiatives which have saved large amounts of money and lives have not led the public to conclude that government initiatives in general are trustworthy and should be expanded, so I'm not convinced that the success of such an initiative would be viewed as a mandate for its expansion.
The costs is an issue of friction; in a vacuum with a spherical Unbreakable Vow, would everyone be bound? Eventually. Why not?
Given the high reported cost, there will be lots of people it's not worth binding, but the exact trade-off will vary. Given the high cost of security and opportunity costs, the cost will have to be large to justify not binding quite a few people (consider how many scores of thousands of dollars it costs to keep an ordinary criminal in prison one year).
Look at what the public does, not what (some of) it says. Governments keep expanding.
Government keeps expanding in some respects, but countries often do not rush to implement programs even when they've proven effective in other countries.
I come back here and I find that gwern has made some of the points I wanted to make, and some point even beyond that. As gwern point out, programs similar in goal and expense although lesser in scope exist in the world today. They have been implemented, so saying that in some (or even most) circumstances countries don't implement this kind of program, is, if not a flawed argument, at least an incomplete one.
Remember how earlier in HPMOR (chapter 47) Harry swore to take as an enemy whoever it was that killed Narcissa Malfoy? It was no an unbreakable vow, but the same principle applies. Not only was he very careful, with many conditions laid upon the pledge, but the first condition said that Draco could release him from the pledge at any time. There is no danger of a vow like that being not perfectly thought out, because if something goes wrong, you can just have whoever you swore it too annul it. I understand that getting a perfect wording is not trivial, but if you just keep a human in the loop like that, you can avoid most errors.
And a clarification:
In a certain sense I am talking about them, as the whole thing started from a discussion of what would happen if we used a specific method to get around the disadvantage, but in practice I am not really talking about them, as with the downside basically gone, there is no real difference between them and an Unbreakable Vow like in canon, with no downside. In other words, I am talking about Unbreakable Vows with no downside, but that could either be the ones we were talking about (with the downside, but with it overcome) or the simpler version which does not have a downside to start out with, and it does not really matter which.
Harry is one of the most intelligent and rational people in the world, and took great care in designing that oath, (which, as you point out, is not unbreakable,) and he's still in a position for it to screw him over, since if Draco's father has been doing his best to change his son's sympathies, then Draco may not be inclined to release Harry from the Vow even if it turns out Dumbledore burned his mother for good reasons.
If Harry had taken an Unbreakable Vow, then even with the escape clause, he would probably be obligated to treat Dumbledore as his enemy right now, with no way to get Draco to release him from it.
There's plenty of danger in an imperfectly thought out vow, even if you add a clause that someone can release you from it. Having someone who could release you from your vow isn't much help if you're already dead due to having been unable to act in self defense, for instance. Supposing you have to go down to the equivalent of a local police station to get released from a Vow, I would suggest that this probably retains most of the problems of being unable to break the vows at all.
If you think it doesn't matter which, I have to suspect that you're not thinking very hard of the implications of the MoR method.
Not everyone can be easily imperiused, nor is everyone capable of casting the spell, and it is probably impossible for a single person to keep a large number of people imperiused at once (canon doesn't say whether it's possible to imperius more than one person at a time, but provides no evidence that it is, and if it were, we could expect people like Voldemort to make extensive use of this.)
If the people being used as binders are not controlled perpetually, then we have a segment of the population which is being victimized in what many humans would regard as one of the most abhorrent ways possible, being routinely mind controlled into performing acts to which they would not consent of their own volition. These people, to put it lightly, do not like the segment of the population which is doing this to them. The people exploiting them need to make arrangements to keep them safely under control, as with chattel slavery. If such arrangements aren't strong enough, they're likely to engage in violent uprisings (although unlikely to succeed if and when they do, they're less powerful and less well coordinated.)
Not only does the society have to invest labor and resources in keeping this segment of the population under wraps, any enemies who want to destabilize this society would do well to target this system. Kind of like the helot system, which was convenient for the Spartans in terms of productivity and military strength, except for the fact that any time they stayed away from home for too long, they were in danger of a revolt, and they had to make all the other city states they dealt with swear to support the system, since it would be so dangerous for them if anyone tried to destabilize it.
You can increase the security by keeping the people shut away somewhere, but then you lose the productivity of the people being used as binders.
Probably the most efficient method would be to force all the binders into unbreakable oaths not to rebel. You are, of course, still losing a significant portion of the total magical powers of the population by using unbreakable vows en-masse like this, and far from other countries seeing and wanting to copy this system, they're liable to see it as either an exploitable weakness or a human rights violation, in which case this society could be facing trade sanctions, embargoes, or even war.
And of course, you still have the issue of how society undergoes the steps to reach this point.
Look at how effective those are in the real world. You have countries ignoring sanctions and embargoes because there's a lot of money to be made that way. As for wars with large coalitions, you have the inevitable issues of members suspecting other members of not holding up their end of the war, or using the war to unfairly increase their power vis-a-vis the other members of the coalition.
Of course, it's not hard to solve all these problems using unbreakable vows, but well. ;)
You are right, Harry did not add enough layers of precautions. As such, he is in a position for it to screw him over. A truly well thought out Vow would have several escape clauses like this, to different people, and also a clause temporarily suspending the Vow while you go to have it removed if you truly believe that the situation warrants its removal.
Even that might not be enough, but remember: Harry thought his oath out in less than a minute. I thought out my additions in not much more. I imagine that if someone smart brainstormed this for a couple of hours, they would figure out even more elaborate and foolproof mechanisms. And once it gets going, there will be people who spend their whole careers on the question, refining the answer even further.
I think we have a very large difference of opinions here. Remember that in the real world, most societies with slaves lasted quite a long while. Add onto that several additional factors that are greatly to the advantage of the masters and not to the advantage of the slaves, and you see why I think there is not much of a difference.
Numbers. In the example you gave, of the Helot system, please not that there were seven times as many helots as non-helots. That's right, seven times as many. In this system, you would not need even nearly a one-to one relationship. I would guess that there would be probably no more then one slave needed for every five or six people. This reduces their ability to revolt by so much that it is nearly impossible to compare it.
Magic. I would envision the slaves not having access to wands except when binding a Vow, giving access to wands to only a small fraction of the slave population at a time. Add on to that the fact the the slaves would not get much magical training, and that they have their magic reserves permanently depleted, and you will see that the masters have another insurmountable advantage.
The Imperius Curse. During the time when the slaves do have a wand, they would be under the Imperius Curse. This would not need more than one slave to be controlled by someone's Imperius Curse at any one point. And remember, when you say that not everyone can be controlled by this curse: In canon, it was a huge surprise that Harry managed to throw it off so easily, the kind of thing you can only get if you are the hero of the story. Most people resistant to it are only that way due to a large amount of willpower built up over the years. And you can just execute people like that.
Unbreakable Vows. I don't think these are needed, as the other factors cover the possibility of a revolt very well, but if needed, some slaves could be forced into an unbreakable vow not to rebel.
Remember that the concept of "human rights violations" are something very new, and they did not stop basically every nation that ever existed from keeping some sort sort of slaves. Remember that for thousands of years, there where, sadly, a lot of slaveholding and other human rights violations. And in no case did this result in the moral outrage from the surrounding countries, and certainly not to the point where they declared war or started trade embargos. (By the way, just thinking about it makes me very much appreciate the moral progress humanity has made in the last few hundred years. It truly is incredible if you think about it.)
We have discussed this a lot, and have provided a lot of explanation about how I do not really consider this an issue, and others have chimed in, adding their views on the subject. After all that, I believe we have provided many different angles of the explanation of why this is. If there is a specific part of it that you object to or want clarified, I am happy to discuss it further, but other than that I am not sure that there is more to discuss.