Perhaps the folks at LW can help me clarify my own conflicting opinions on a matter I’ve been giving a bit of thought lately.
Until about the time I left for college, most of my views reflected those of my parents. It was a pretty common Republican party-line cluster, and I’ve got concerns that I have anchored at a point too close to favoring the death penalty than I should. I read studies about how capital punishment disproportionately harms minorities, and I think Robin Hanson had more to say about difference in social tier. Early in my college time, this sort of problem led me to reject the death penalty on practical grounds. Then, as I lost my religious views, I stopped seeing it as a punishment at all. I started to see it as a the same basic thing as putting down an aggressive dog. After all, dead people have a pretty encouraging recidivism rate.
I began to wonder if I could reject the death penalty on principle. A large swath of America believes that the words of the Declaration of Independence are as pertinent to our country as the Constitution. This would mean that we could disallow execution because it conflicts with our “inalienable” right to life. But then, I can’t justify using the same argument as the people who try to prove that America is a Christian nation. As an interesting corollary, it seems that anyone citing the Declaration in this manner will have a very hard time also supporting the death penalty for this reason.
So basically, I think I would find the death penalty morally acceptable, but only in the hypothetical realm of virtual certainty that the inmate is guilty of a heinous crime. And I have no bound for what that virtual certainty is. Certainly a 5% chance of being falsely accused is too high. I wouldn’t kill one innocent man to rid the world of 19 bad ones. But then, I would kill an innocent person to stop a billion headaches (an example I just read in Steven Landsburg’s The Big Questions), so I obviously don’t demand 100% certainty.
It seems like I might be asking: “What are the chances that someone was falsely accused, given that they were accused of an execution-worthy crime?” And a follow-up “What is an acceptable chance for killing an innocent person?”
Can Bayes help here? I am eager to hear some actual opinions on this matter. So far I’ve come up with precious little when talking to friends and family.
My take on capital punishment is that it’s not actually that important an issue. With pretty much anything that you can say about the death penalty, you can say something similar about life imprisonment without parole (especially with the way that the death penalty is actually practiced in the United States). Would you lock an innocent man in a cell for the rest of his life to keep 19 bad ones locked up?
Virtually zero chance of recidivism? True for both. Very expensive? Check. Wrongly convicted innocent people get screwed? Check—though in both cases they have a decent chance of being exonerated after conviction before getting totally screwed (and thus only being partially screwed). Could be considered immoral to do something so severe to a person? Check. Deprives people of an “inalienable” right? Check (life/liberty). Strongly demonstrates society’s disapproval of a crime? Check (slight edge to capital punishment, though life sentences would be better at this if the death penalty wasn’t an option). Applied disproportionately to certain groups? I think so, though I don’t know the research. Strong deterrent? It seems like the death penalty should be a bit stronger, but the evidence is unclear on that. Provides closure to the victim’s family? Execution seems like more definitive closure, but they have to wait until years after sentencing to get it.
The criminal justice system is a big important topic, and I think it’s too bad that this little piece of it (capital punishment) soaks up so much of our attention to it. Overall, my stance on capital punishment is ambivalent, leaning against it because it’s not worth the trouble, though in some cases (like McVeigh) it’s nice to have around and I could be swayed by a big deterrent effect. I’d prefer for more of the focus to be on this sort of thing (pdf).
Good post. I have never seen strong evidence that the death penalty has a meaningful deterrent effect but I’d be curious to see links one way or the other.
I lean towards prison abolition, but it’s an idealistic notion, not a pragmatic one. I suppose we could start by getting rid of prisons for non-violent crimes and properly funding mental hospitals. http://en.wikipedia.org/wiki/Prison_abolition_movement I can’t see that happening when we can’t even decriminalize marijuana.
Personal response: I’m opposed to the death penalty because it costs more than putting them in prison for life due to the huge number of appeals they’re allowed (vaguely recall hearing in newspapers / reports). I feel the US has become so risk-averse and egalitarian that it cannot properly implement a death penalty. This is reflected in the back-and-forth questions you ask.
I also oppose it on the grounds that it is often used as a tool of vengeance rather than justice. Nitrogen poisoning (I think that was the gas they were talking about) is a safe, highly reliable, and euphoric means of death, but the US still prefers electrocution (can take minutes), injection (can feel like the veins are burning from the inside out while the body is paralyzed), etc.
That said, I don’t care enough about the topic to try and alter its use, whether through voting, polling, letters, etc, nor do I desire to put much thought into it. Best to let hot topics alone.
And after asking about Bayes, you should ask for math rather than opinions.
There is strong Bayesian evidence that the USA has executed one innocent man. http://en.wikipedia.org/wiki/Cameron_Todd_Willingham By that I mean that an Amanda Knox test type analysis would clearly show that Willingham is innocent, probably with greater certainty than when the Amanda Knox case was analyzed. Does knowing that the USA has indeed provably executed an innocent person change your opinion?
What are the practical advantages of death over life in prison? US law allows for true life without parole. Life in an isolated cell in a Supermax prison is continual torture—it is not a light punishment by any means. Without a single advantage given for the death penalty over life in prison without parole, I think that ~100% certainty is needed for execution.
I am against the death penalty for regular murder and mass murder and aggravated rape. I am indifferent with regards to the death penalty for crimes against humanity as I recognize that symbolic execution could be appropriate for grave enough crimes.
Kevin, thank you for the specific example. It definitely strengthened my practical objection to the practice. I strongly suspect that the current number of false positives lies outside of my acceptance zone.
Rain, I agree that politics is a mind-killer, but thought it worthy of at least brushing the cobwebs off some cached thoughts. Good point about Nitrogen. I wonder why we choose gruesome methods when even CO would be cheap, easy and effective.
Morendil, I appreciate the other questions. You have a good point that if Omega were brought in on the justice system, it would definitely find better corrective measures than the kill command. I think Eliezer once talked about how predicting your possible future decisions is basically the same as deciding. In that case, I already changed many things on this Big Question, and am just finally doing what I predicted I might do last time I gave any thought to capital punishment. Which happened to be at the conclusion (if there is such a thing) of a murder trial where my friend was a victim. Lots of bias to overcome there, methinks.
Unnamed, interesting points. I hadn’t actually considered how similar life imprisonment is to execution, with regard to the pertinent facts. I was recently introduced to the concept of restorative justice which I think encompasses your article. I find it particularly appealing because it deals with what works, instead of worthless Calvinist ideals like punishment. From my understanding, execution only fulfills punishment in the most trivial of senses.
I am against the death penalty for regular murder and mass murder and aggravated rape. I am indifferent with regards to the death penalty for crimes against humanity as I recognize that symbolic execution could be appropriate for grave enough crimes.
“Crimes against humanity” is one of the crimes that for most practical purposes means ”… and lost”.
Yup. Even though they’ll never face charges, some of the winners are guilty as sin. And I mean that the Project for the New American Century was on the winning side of the war, their namesake mission has failed horribly.
The more judicious question, I am coming to realize, isn’t so much “Which of these two Standard Positions should I stand firmly on”.
The more useful question is, why do the positions matter? Why is the discussion currently crystallized around these standard positions important to me, and how should I fluidly allow whatever evidence I can find to move me toward some position, which is rather unlikely (given that the debate has been so long crystallized in this particular way) to be among the standard ones. And I shouldn’t necessarily expect to stay at that position forever, once I have admitted in principle that new evidence, or changes in other beliefs of mine, must commit me to a change in position on that particular issue.
In the death-penalty debate I identify more strongly with the “abolitionist” standard position because I was brought up in an abolitionist country by left-wing parents. That is, I find myself on the opposite end of the spectrum from you. And yet, perhaps we are closer than is apparent at first glance, if we are both of us committed primarily to investigating the questions of values, the questions of fact, and the questions of process that might leave either or both of us, at the end of the inquiry, in a different position than we started from.
Would I revise my “in principle” opposition to the death penalty if, for instance, the means of “execution” were modified to cryonic preservation? Would I then support cryonic preservation as a “punishment” for lesser crimes such as would currently result in lifetime imprisonment?
Would I still oppose the death penalty if we had a Truth Machine? Or if we could press Omega into service to give us a negligible probability of wrongful conviction? Or otherwise rely on a (putatively) impartial means of judgment which didn’t involve fallible humans? Is that even desirable, if it was at all possible?
Would I support the death penalty if I found out it was an effective deterrent, or would I oppose it only if I found that it didn’t deter? Does deterrence matter? Why, or why not?
How does economics enter into such a decision? How much, whatever position I arrive at, should I consider myself obligated to actively try to ensure that the society I live in espouses that position? For what scope of “the society I live in”—how local or global?
Those are topics and questions I encounter in the process of thinking about things other than the death penalty; practically every important topic has repercussions on this one.
There’s an old systems science saying that I think applies to rational discussions about Big Questions such as this one: “you can’t change just one thing”. You can’t decide on just one belief, and as I have argued before, it serves no useful purpose to call an isolated belief “irrational”. It seems more appropriate to examine the processes whereby we adjust networks of beliefs, how thoroughly we propagate evidence and argument among those networks.
There is currently something of a meta-debate on LW regarding how best to reflect this networked structure of adjusting our beliefs based on evidence and reasoning, with approaches such as TakeOnIt competing against more individual debate modeling tools, with LessWrong itself, not so much the blog but perhaps the community and its norms, having some potential to serve as such a process for arbitrating claims.
But all these prior discussions seem to take as a starting point that “you can’t change just one belief”. That’s among the consequences of embracing uncertainty, I think.
Well, even relatively uncontroversial topics have the same entangled-with-your-entire-belief-network quality to them, but (to most people) less power to make you care.
The judicious response to that is to exercise some prudence in the things you choose to care about. If you care too much about things you have little power to influence and could easily be wrong about, you end up “mind-killed”. If you care too little and about too few things except for basic survival, you end up living the kind of life where it makes little difference how rational you are.
The way it’s worked out for me is that I’ve lived through some events which made me feel outraged, and for better or for worse the outrage made me care about some particular topics, and caring about these topics has made me want to be right about them. Not just to associate myself with the majority, or with a set of people I’d pre-determined to be “the right camp to be in”, but to actually be right.
Political questions like this are far removed from the kind of analysis you seem to want to apply. If it’s you taking out a killer yourself that’s one thing, but the question of whether to support it as a law is something entirely different. This rabbit hole goes very far indeed. Anyway, why would you care about the Constitution—you’re not one of the signers, are you? ;-)
I care about the Constitution for a couple of reasons beyond the narrowly patriotic:
(1) For the framers, its design posed a problem very similar to the design of Friendly AI. The newly independent British colonies were in a unique situation. On the one hand, whatever sort of nation they designed was likely to become quite powerful; it had good access to very large quantities of people, natural resources, and ideas, and the general culture of empiricism and liberty meant that the nation behaved as if it were much more intelligent than most of its competitors. On the other hand, the design they chose for the government that would steer that nation was likely to be quite permanent; it is one thing to change your system of government as you are breaking away from a distant and unpopular metropole, and another to change your government once that government is locally rooted and supported. The latter takes a lot more blood, and carries a much higher risk of simply descending into medium-term anarchy. Finally, the Founders knew that they could not see every possible obstacle that the young and unusual nation would encounter, and so they would have to create a system that could learn based on input from its environment without further input from its designers. So just as we have to figure out how to design a system that will usefully manage vast resources and intelligence in situations we cannot fully predict and with directions that, once issued, cannot be edited or recalled, so too did the Founding Fathers, and we should try to learn from their failures and successes.
(2) The Constitution has come to embody, however imperfectly, some of the core tenents of Bayesianism. I quote Chief Justice Oliver Wendell Holmes:
Persecution for the expression of opinions seems to me perfectly logical. If you have no doubt of your premises or your power and want a certain result with all your heart you naturally express your wishes in law and sweep away all opposition...But when men have realized that time has upset many fighting faiths, they may come to believe even more than they believe the very foundations of their own conduct that the ultimate good desired is better reached by free trade in ideas...that the best test of truth is the power of the thought to get itself accepted in the competition of the market, and that truth is the only ground upon which their wishes safely can be carried out. That at any rate is the theory of our Constitution.
Re 1, if that is the case why not support the Articles of Confederation instead? I also take exception to the underlying assumption that society needs top-down designing, but that’s a very deep debate.
But when men have realized that time has upset many fighting faiths, they may come to believe even more than they believe the very foundations of their own conduct that the ultimate good desired is better reached by free trade in ideas. … hat at any rate is the theory of our Constitution.
If that was really the theory—“checks and balances”—the Constitution was a huge step backward from the Articles of Confederation. (I don’t support the AoC, but I’d prefer them to the Constitution.)
Re 1, if that is the case why not support the Articles of Confederation instead?
I never said we should support it; I said we should care about it.
It would be silly to claim that anyone interested in FAI should be pro-Constitution; there were plenty of 18th century people who earnestly grappled with their version of the FAI problem and thought the Constitution was a bad idea. If you agree more with the anti-Federalists, fine! The point is that we should closely follow the results of the experiment, not that we should bark agreement with the particular set of hypotheses chosen by the Founding Fathers for extensive testing.
Very good point, and the founders’ process for developing the constitution and bill of rights is important for thinking about how to develop a Friendly (mostly Friendly?) AI.
Anyway, why would you care about the Constitution—you’re not one of the signers, are you? ;-)
I swore an oath to support and defend the Constitution as a condition of employment, so at the very least I have to signal caring about it. I doubt beriukay is in the same position, though.
Oaths in general can be a form of precommitment and a weak signal that someone ascribes to certain moral or legal values, though no one seemed to take it seriously in this instance. On my first day, it was just another piece of paper in with all the other forms they wanted me to sign, and they took it away right after a perfunctory reading. I had to search it out online to remember just what it was I had sworn to do. Later, I learned some people didn’t even remember they had taken it.
Personally, I consider it very important to know the rules, laws, commitments, etc., for which I may be responsible, so when I or someone else breaks them, I can clearly note it.
For example, in middle school, one of my teachers didn’t like me whispering to the person sitting next to me in class. When she asked what I was doing, I told her that I was explaining the lesson, since she did a poor job of it. She asked me if I would like to be suspended for disrespect; I made sure to let her know that the form did not have ‘disrespect’ as a reason for suspension, only detention.
Personally, I consider it very important to know the rules, laws, commitments, etc., for which I may be responsible, so when I or someone else breaks them, I can clearly note it.
Far out. That is important.
As for your story, it’s something I would have done but I hope you understand that a little tact could have gone a long way.
What I was trying to get at you seem to think also. You think you are sending a ‘weak signal’ that you are committed to something. But you are using words that I think many around here would be suspicious of (e.g. oath and sworn).
You can say you will do something. If someone doesn’t trust that assertion, how will they ever trust ‘no really I’m serious’.
You can say you will do something. If someone doesn’t trust that assertion, how will they ever trust ‘no really I’m serious’.
Perhaps through enforcement. There are a significant number of laws, regulations, and directives that cover US Federal employees, and the oath I linked to above is a signed and sworn statement indicating the fact that I am aware of and accept responsibility for them.
She asked me if I would like to be suspended for disrespect; I made sure to let her know that the form did not have ‘disrespect’ as a reason for suspension, only detention.
You prefer more time locked up in school than less?
My explanation: It is ironic that ‘more time at school after it finishes’ is used as a punishment and yet ‘days off school’ is considered a worse punishment.
Given the chance I would go back in time and explain to my younger self that just because something is presented as a punishment or ‘worse punishment’ doesn’t mean you have to prefer to avoid it. Further, I would explain that getting what he wants does not always require following the rules presented to him. He can make his own rules, chose among preferred consequences.
While I never actually got either a detention or a suspension, I would have to say I’d prefer the suspension.
In theory but I wonder how long it has been since you were in school. In GA they got around to making a rule that if you were suspended you would lose your drivers license. Also, suspensions typically imply a 0 on all assignments (and possibly tests) that were due for its duration.
My explanation: It is ironic that ‘more time at school after it finishes’ is used as a punishment and yet ‘days off school’ is considered a worse punishment.
Given the chance I would go back in time and explain to my younger self that just because something is presented as a punishment or ‘worse punishment’ doesn’t mean you have to prefer to avoid it. Further, I would explain that getting what he wants does not always require following the rules presented to him. He can make his own rules, chose among preferred consequences.
While I never actually got either a detention or a suspension, I would have to say I’d prefer the suspension.
Hailing from secular Britain I wasn’t aware of the distinction. Affirmation actually sounds more religious to me. I’d never particularly associated the idea of an oath with religion but I can see how such an association could sour one on the word ‘oath’.
Well obviously the idea of an oath only has value if it is credible, that is why there are often strong cultural taboos against oath breaking. In times past there were often harsh punishments for oath breaking to provide additional enforcement but it is true that in the modern world much of the function of oaths has been transferred to the legal system. Traditionally one of the things that defined a profession was the expectation that members of the profession held themselves to a standard above and beyond the minimum enforced by law however. Professional oaths are part of that tradition, as is the idea of an oath sworn by civil servants and other government employees. This general concept is not unique to the US or to government workers.
Perhaps the folks at LW can help me clarify my own conflicting opinions on a matter I’ve been giving a bit of thought lately.
Until about the time I left for college, most of my views reflected those of my parents. It was a pretty common Republican party-line cluster, and I’ve got concerns that I have anchored at a point too close to favoring the death penalty than I should. I read studies about how capital punishment disproportionately harms minorities, and I think Robin Hanson had more to say about difference in social tier. Early in my college time, this sort of problem led me to reject the death penalty on practical grounds. Then, as I lost my religious views, I stopped seeing it as a punishment at all. I started to see it as a the same basic thing as putting down an aggressive dog. After all, dead people have a pretty encouraging recidivism rate.
I began to wonder if I could reject the death penalty on principle. A large swath of America believes that the words of the Declaration of Independence are as pertinent to our country as the Constitution. This would mean that we could disallow execution because it conflicts with our “inalienable” right to life. But then, I can’t justify using the same argument as the people who try to prove that America is a Christian nation. As an interesting corollary, it seems that anyone citing the Declaration in this manner will have a very hard time also supporting the death penalty for this reason.
So basically, I think I would find the death penalty morally acceptable, but only in the hypothetical realm of virtual certainty that the inmate is guilty of a heinous crime. And I have no bound for what that virtual certainty is. Certainly a 5% chance of being falsely accused is too high. I wouldn’t kill one innocent man to rid the world of 19 bad ones. But then, I would kill an innocent person to stop a billion headaches (an example I just read in Steven Landsburg’s The Big Questions), so I obviously don’t demand 100% certainty.
It seems like I might be asking: “What are the chances that someone was falsely accused, given that they were accused of an execution-worthy crime?” And a follow-up “What is an acceptable chance for killing an innocent person?”
Can Bayes help here? I am eager to hear some actual opinions on this matter. So far I’ve come up with precious little when talking to friends and family.
My take on capital punishment is that it’s not actually that important an issue. With pretty much anything that you can say about the death penalty, you can say something similar about life imprisonment without parole (especially with the way that the death penalty is actually practiced in the United States). Would you lock an innocent man in a cell for the rest of his life to keep 19 bad ones locked up?
Virtually zero chance of recidivism? True for both. Very expensive? Check. Wrongly convicted innocent people get screwed? Check—though in both cases they have a decent chance of being exonerated after conviction before getting totally screwed (and thus only being partially screwed). Could be considered immoral to do something so severe to a person? Check. Deprives people of an “inalienable” right? Check (life/liberty). Strongly demonstrates society’s disapproval of a crime? Check (slight edge to capital punishment, though life sentences would be better at this if the death penalty wasn’t an option). Applied disproportionately to certain groups? I think so, though I don’t know the research. Strong deterrent? It seems like the death penalty should be a bit stronger, but the evidence is unclear on that. Provides closure to the victim’s family? Execution seems like more definitive closure, but they have to wait until years after sentencing to get it.
The criminal justice system is a big important topic, and I think it’s too bad that this little piece of it (capital punishment) soaks up so much of our attention to it. Overall, my stance on capital punishment is ambivalent, leaning against it because it’s not worth the trouble, though in some cases (like McVeigh) it’s nice to have around and I could be swayed by a big deterrent effect. I’d prefer for more of the focus to be on this sort of thing (pdf).
Good post. I have never seen strong evidence that the death penalty has a meaningful deterrent effect but I’d be curious to see links one way or the other.
I lean towards prison abolition, but it’s an idealistic notion, not a pragmatic one. I suppose we could start by getting rid of prisons for non-violent crimes and properly funding mental hospitals. http://en.wikipedia.org/wiki/Prison_abolition_movement I can’t see that happening when we can’t even decriminalize marijuana.
Standard response: politics is the mind-killer.
Personal response: I’m opposed to the death penalty because it costs more than putting them in prison for life due to the huge number of appeals they’re allowed (vaguely recall hearing in newspapers / reports). I feel the US has become so risk-averse and egalitarian that it cannot properly implement a death penalty. This is reflected in the back-and-forth questions you ask.
I also oppose it on the grounds that it is often used as a tool of vengeance rather than justice. Nitrogen poisoning (I think that was the gas they were talking about) is a safe, highly reliable, and euphoric means of death, but the US still prefers electrocution (can take minutes), injection (can feel like the veins are burning from the inside out while the body is paralyzed), etc.
That said, I don’t care enough about the topic to try and alter its use, whether through voting, polling, letters, etc, nor do I desire to put much thought into it. Best to let hot topics alone.
And after asking about Bayes, you should ask for math rather than opinions.
Yeah, my formatting of the last few sentences wasn’t very great. Sorry.
There is strong Bayesian evidence that the USA has executed one innocent man. http://en.wikipedia.org/wiki/Cameron_Todd_Willingham By that I mean that an Amanda Knox test type analysis would clearly show that Willingham is innocent, probably with greater certainty than when the Amanda Knox case was analyzed. Does knowing that the USA has indeed provably executed an innocent person change your opinion?
What are the practical advantages of death over life in prison? US law allows for true life without parole. Life in an isolated cell in a Supermax prison is continual torture—it is not a light punishment by any means. Without a single advantage given for the death penalty over life in prison without parole, I think that ~100% certainty is needed for execution.
I am against the death penalty for regular murder and mass murder and aggravated rape. I am indifferent with regards to the death penalty for crimes against humanity as I recognize that symbolic execution could be appropriate for grave enough crimes.
Kevin, thank you for the specific example. It definitely strengthened my practical objection to the practice. I strongly suspect that the current number of false positives lies outside of my acceptance zone.
Rain, I agree that politics is a mind-killer, but thought it worthy of at least brushing the cobwebs off some cached thoughts. Good point about Nitrogen. I wonder why we choose gruesome methods when even CO would be cheap, easy and effective.
Morendil, I appreciate the other questions. You have a good point that if Omega were brought in on the justice system, it would definitely find better corrective measures than the kill command. I think Eliezer once talked about how predicting your possible future decisions is basically the same as deciding. In that case, I already changed many things on this Big Question, and am just finally doing what I predicted I might do last time I gave any thought to capital punishment. Which happened to be at the conclusion (if there is such a thing) of a murder trial where my friend was a victim. Lots of bias to overcome there, methinks.
Unnamed, interesting points. I hadn’t actually considered how similar life imprisonment is to execution, with regard to the pertinent facts. I was recently introduced to the concept of restorative justice which I think encompasses your article. I find it particularly appealing because it deals with what works, instead of worthless Calvinist ideals like punishment. From my understanding, execution only fulfills punishment in the most trivial of senses.
“Crimes against humanity” is one of the crimes that for most practical purposes means ”… and lost”.
Yup. Even though they’ll never face charges, some of the winners are guilty as sin. And I mean that the Project for the New American Century was on the winning side of the war, their namesake mission has failed horribly.
The more judicious question, I am coming to realize, isn’t so much “Which of these two Standard Positions should I stand firmly on”.
The more useful question is, why do the positions matter? Why is the discussion currently crystallized around these standard positions important to me, and how should I fluidly allow whatever evidence I can find to move me toward some position, which is rather unlikely (given that the debate has been so long crystallized in this particular way) to be among the standard ones. And I shouldn’t necessarily expect to stay at that position forever, once I have admitted in principle that new evidence, or changes in other beliefs of mine, must commit me to a change in position on that particular issue.
In the death-penalty debate I identify more strongly with the “abolitionist” standard position because I was brought up in an abolitionist country by left-wing parents. That is, I find myself on the opposite end of the spectrum from you. And yet, perhaps we are closer than is apparent at first glance, if we are both of us committed primarily to investigating the questions of values, the questions of fact, and the questions of process that might leave either or both of us, at the end of the inquiry, in a different position than we started from.
Would I revise my “in principle” opposition to the death penalty if, for instance, the means of “execution” were modified to cryonic preservation? Would I then support cryonic preservation as a “punishment” for lesser crimes such as would currently result in lifetime imprisonment?
Would I still oppose the death penalty if we had a Truth Machine? Or if we could press Omega into service to give us a negligible probability of wrongful conviction? Or otherwise rely on a (putatively) impartial means of judgment which didn’t involve fallible humans? Is that even desirable, if it was at all possible?
Would I support the death penalty if I found out it was an effective deterrent, or would I oppose it only if I found that it didn’t deter? Does deterrence matter? Why, or why not?
How does economics enter into such a decision? How much, whatever position I arrive at, should I consider myself obligated to actively try to ensure that the society I live in espouses that position? For what scope of “the society I live in”—how local or global?
Those are topics and questions I encounter in the process of thinking about things other than the death penalty; practically every important topic has repercussions on this one.
There’s an old systems science saying that I think applies to rational discussions about Big Questions such as this one: “you can’t change just one thing”. You can’t decide on just one belief, and as I have argued before, it serves no useful purpose to call an isolated belief “irrational”. It seems more appropriate to examine the processes whereby we adjust networks of beliefs, how thoroughly we propagate evidence and argument among those networks.
There is currently something of a meta-debate on LW regarding how best to reflect this networked structure of adjusting our beliefs based on evidence and reasoning, with approaches such as TakeOnIt competing against more individual debate modeling tools, with LessWrong itself, not so much the blog but perhaps the community and its norms, having some potential to serve as such a process for arbitrating claims.
But all these prior discussions seem to take as a starting point that “you can’t change just one belief”. That’s among the consequences of embracing uncertainty, I think.
Yeah, that’s why I try to avoid hot topics. Too much work.
Well, even relatively uncontroversial topics have the same entangled-with-your-entire-belief-network quality to them, but (to most people) less power to make you care.
The judicious response to that is to exercise some prudence in the things you choose to care about. If you care too much about things you have little power to influence and could easily be wrong about, you end up “mind-killed”. If you care too little and about too few things except for basic survival, you end up living the kind of life where it makes little difference how rational you are.
The way it’s worked out for me is that I’ve lived through some events which made me feel outraged, and for better or for worse the outrage made me care about some particular topics, and caring about these topics has made me want to be right about them. Not just to associate myself with the majority, or with a set of people I’d pre-determined to be “the right camp to be in”, but to actually be right.
Political questions like this are far removed from the kind of analysis you seem to want to apply. If it’s you taking out a killer yourself that’s one thing, but the question of whether to support it as a law is something entirely different. This rabbit hole goes very far indeed. Anyway, why would you care about the Constitution—you’re not one of the signers, are you? ;-)
I care about the Constitution for a couple of reasons beyond the narrowly patriotic:
(1) For the framers, its design posed a problem very similar to the design of Friendly AI. The newly independent British colonies were in a unique situation. On the one hand, whatever sort of nation they designed was likely to become quite powerful; it had good access to very large quantities of people, natural resources, and ideas, and the general culture of empiricism and liberty meant that the nation behaved as if it were much more intelligent than most of its competitors. On the other hand, the design they chose for the government that would steer that nation was likely to be quite permanent; it is one thing to change your system of government as you are breaking away from a distant and unpopular metropole, and another to change your government once that government is locally rooted and supported. The latter takes a lot more blood, and carries a much higher risk of simply descending into medium-term anarchy. Finally, the Founders knew that they could not see every possible obstacle that the young and unusual nation would encounter, and so they would have to create a system that could learn based on input from its environment without further input from its designers. So just as we have to figure out how to design a system that will usefully manage vast resources and intelligence in situations we cannot fully predict and with directions that, once issued, cannot be edited or recalled, so too did the Founding Fathers, and we should try to learn from their failures and successes.
(2) The Constitution has come to embody, however imperfectly, some of the core tenents of Bayesianism. I quote Chief Justice Oliver Wendell Holmes:
Re 1, if that is the case why not support the Articles of Confederation instead? I also take exception to the underlying assumption that society needs top-down designing, but that’s a very deep debate.
If that was really the theory—“checks and balances”—the Constitution was a huge step backward from the Articles of Confederation. (I don’t support the AoC, but I’d prefer them to the Constitution.)
I never said we should support it; I said we should care about it.
It would be silly to claim that anyone interested in FAI should be pro-Constitution; there were plenty of 18th century people who earnestly grappled with their version of the FAI problem and thought the Constitution was a bad idea. If you agree more with the anti-Federalists, fine! The point is that we should closely follow the results of the experiment, not that we should bark agreement with the particular set of hypotheses chosen by the Founding Fathers for extensive testing.
Very good point, and the founders’ process for developing the constitution and bill of rights is important for thinking about how to develop a Friendly (mostly Friendly?) AI.
I swore an oath to support and defend the Constitution as a condition of employment, so at the very least I have to signal caring about it. I doubt beriukay is in the same position, though.
Do you really take that sort of thing seriously? Far out if you do, but I have trouble with the concept of an ‘oath’.
Oaths in general can be a form of precommitment and a weak signal that someone ascribes to certain moral or legal values, though no one seemed to take it seriously in this instance. On my first day, it was just another piece of paper in with all the other forms they wanted me to sign, and they took it away right after a perfunctory reading. I had to search it out online to remember just what it was I had sworn to do. Later, I learned some people didn’t even remember they had taken it.
Personally, I consider it very important to know the rules, laws, commitments, etc., for which I may be responsible, so when I or someone else breaks them, I can clearly note it.
For example, in middle school, one of my teachers didn’t like me whispering to the person sitting next to me in class. When she asked what I was doing, I told her that I was explaining the lesson, since she did a poor job of it. She asked me if I would like to be suspended for disrespect; I made sure to let her know that the form did not have ‘disrespect’ as a reason for suspension, only detention.
Far out. That is important.
As for your story, it’s something I would have done but I hope you understand that a little tact could have gone a long way.
What I was trying to get at you seem to think also. You think you are sending a ‘weak signal’ that you are committed to something. But you are using words that I think many around here would be suspicious of (e.g. oath and sworn).
You can say you will do something. If someone doesn’t trust that assertion, how will they ever trust ‘no really I’m serious’.
Perhaps through enforcement. There are a significant number of laws, regulations, and directives that cover US Federal employees, and the oath I linked to above is a signed and sworn statement indicating the fact that I am aware of and accept responsibility for them.
You prefer more time locked up in school than less?
No.
My explanation: It is ironic that ‘more time at school after it finishes’ is used as a punishment and yet ‘days off school’ is considered a worse punishment.
Given the chance I would go back in time and explain to my younger self that just because something is presented as a punishment or ‘worse punishment’ doesn’t mean you have to prefer to avoid it. Further, I would explain that getting what he wants does not always require following the rules presented to him. He can make his own rules, chose among preferred consequences.
While I never actually got either a detention or a suspension, I would have to say I’d prefer the suspension.
In theory but I wonder how long it has been since you were in school. In GA they got around to making a rule that if you were suspended you would lose your drivers license. Also, suspensions typically imply a 0 on all assignments (and possibly tests) that were due for its duration.
As a teacher or a student? 4 years and respectively.
My explanation: It is ironic that ‘more time at school after it finishes’ is used as a punishment and yet ‘days off school’ is considered a worse punishment.
Given the chance I would go back in time and explain to my younger self that just because something is presented as a punishment or ‘worse punishment’ doesn’t mean you have to prefer to avoid it. Further, I would explain that getting what he wants does not always require following the rules presented to him. He can make his own rules, chose among preferred consequences.
While I never actually got either a detention or a suspension, I would have to say I’d prefer the suspension.
I have a martyr complex.
How so?
An oath is an appeal to a sacred witness, typically the God of Abraham. An affirmation is the secular version of an oath in the American legal system.
Hailing from secular Britain I wasn’t aware of the distinction. Affirmation actually sounds more religious to me. I’d never particularly associated the idea of an oath with religion but I can see how such an association could sour one on the word ‘oath’.
Yeah I like Kevin’s short answer. But in general I said to Rain:
When you make something a contract you see there are some legal teeth, but swearing to uphold the constitution feels silly.
Well obviously the idea of an oath only has value if it is credible, that is why there are often strong cultural taboos against oath breaking. In times past there were often harsh punishments for oath breaking to provide additional enforcement but it is true that in the modern world much of the function of oaths has been transferred to the legal system. Traditionally one of the things that defined a profession was the expectation that members of the profession held themselves to a standard above and beyond the minimum enforced by law however. Professional oaths are part of that tradition, as is the idea of an oath sworn by civil servants and other government employees. This general concept is not unique to the US or to government workers.