I’m not sure that was Dominic Cummings’ strategy? AFAICT Dominic Cummings is/was a regular here on LessWrong, from perhaps as far back as the OvercomingBias days, though I don’t know his account handle. Look at his blog post which quotes Eliezer Yudkowsky first, and is chalk-full of LessWrong-defined non-standard vocabulary and phrasing. He’s not (just) advertising to rationalists, that’s how he always communicates.
My understanding is that he was, at minimum, a long-time lurker here who some point he decided, pretty much on his own, that the UK leaving the EU was objectively the best outcome for “reasons”.. and then used all the tricks in the rationalist dark-arts book to make it happen. He was successful at winning, and radically changed the course of a major country.
We talk a big talk about instrumental rationality, about rationalist superpowers… but application has been a bit of a mixed bag to be honest. And yet here’s a rationalist who upturned global politics singlehandedly, and credits LessWrong with his success.
I would suggest setting aside ALL feelings about Brexit from your mind (politics is the mind-killer) and studying from an objective viewpoint what Dominic Cummings has done. There is something in the process there that ought to be emulated, even if you disagree with the instrumental outcome.
Burning down a building is easier than constructing it.
People are celebrating Dominic Cummings for changing the building. I’d like to wait until it turns out what specific kind of change it was.
In the meanwhile, I accept the argument that even burning down the building requires more skills and agency than merely talking about the building. In this way, Dominic Cummings has already risen above the level of the rationalist plebs. But how high, that still remains to be seen.
Don’t confuse the consequences of the outcome with the things that had to be achieved to accomplish it. He got a national referendum on the ballot and got it to pass, despite the establishment (political parties, media, and businesses) being uniformly against it. This is not comparable to burning down a building. Rather it was a precision campaign to identify what it would take to convince a majority of the voting population to adopt a radical political agenda.
Change “Brexit” to “basic income,” “universal single-payer healthcare,” “tuition-free higher education,” “carbon-free emissions” or whatever your preferred legislative objective would be, and the difficulty and techniques would be the same. It was a large accomplishment worth studying.
It’s not obvious to me that either the difficulty or the techniques would be the same for those other objectives as for Brexit. A canny political operative uses techniques appropriate for specific goals, after all.
If your goal is simply to get the UK out of the EU, for whatever reason and in whatever fashion, and if you don’t mind what harm you do to society in the process, then “all” you need to do is to stir up hatred and suspicion and fear around the EU and what it does and those who like it, and find some slogans that appeal without requiring much actual thought, and so forth. Standard-issue populism.
But let’s suppose you want universal basic income and you want it because you think it will help people and make society better. Then:
The “stir up fear and hatred” template doesn’t work so well, because what you’re doing isn’t a thing that can readily be seen as fighting against a shared enemy.
The “stir up fear and hatred” template may be a really bad idea even if it works, because it may do more damage to society than the reform you’re aiming for does good.
The details of what you do and how you do it may matter a lot: some versions of universal basic income might bankrupt the country, some might fail to do enough to solve the problems UBI is meant to solve, some might be politically unacceptable, etc. So you need to sell it in a way that lets a carefully designed version of UBI be what ends up happening.
The available evidence does not suggest (to me) that Cummings has a very specific version of Brexit in mind, or that he is sufficiently concerned for the welfare of the UK’s society and the individuals within it to be troubled by considerations of societal harm done by the measures he’s taken, or of whether he’s ending up with a variety of Brexit that’s net beneficial.
I would have preferred to say the foregoing without the last paragraph, which is kinda object-level political. But it’s essential to the point. When Viliam says it’s easier to burn a building down than construct it, I think he is saying something similar: if, as it seems may be the case, Cummings doesn’t actually care whether he does a lot of harm to a lot of people, then he has selected an easier task than would be faced by someone trying to bring about major reforms without harming a lot of people, and the methods he’s chosen are not necessarily ones that those who care about not harming a lot of people should emulate.
Yep. Looking around me, getting Slovakia out of EU would be relatively easier task than making it adopt UBI, for the reasons you mentioned (plus one you didn’t: availability of foreign helpers).
How clear is it that he specifically got all those things to happen? There were definitely other people involved, after all. Cummings’s own account of what happened no doubt ascribes as much agency as possible to Cummings himself, but there are possible explanations for that other than its being true.
This conversation sounds like TAG uses utilitarianism to mean classic utilitarianism, where pains and pleasures are the only consequences that we care about (and rights violations are not), and like you are using it to refer to decision-theoretic utilitarianism, where the consequences can include rights violations as well.
I object that there is any real difference between the two. Classic utilitarianism is “count up all the good, subtract all the bad.” That’s exactly how I would describe decision-theoretic utilitarianism. Now the original proponents of utilitarianism also didn’t give much negative weight to rights violations, but that’s a complaint against their utility function, not utilitarianism per se.
But regardless, I think you did identify and elucidate better than I did the core disagreement here. I hope it’s resolved for TAG.
It sounds like “decision theoretic utilitarianism” was something invented here.
I think hybrid approaches to ethics have more to offer than purist approaches..and also that it is assists communication to label them as such.
Edit:
Actually , it’s worse than that. As Smiffnoy correctly states, maximising your personal utility without regard to anybody else isn’t an ethical theory at all, so it continues the confusion to label it as such.
Actually , it’s worse than that. As Smiffnoy correctly states, maximising your personal utility without regard to anybody else isn’t an ethical theory at all, so it continues the confusion to label it as such.
That only describes a solipsist or sociopath’s utility function. All things being equal, I would like for you to be happy, strange person on the Internet who is reading this. Maximizing my own utility function means preferring outcomes where everyone is happy, because I value those outcomes.
Also Smiffnoy seems to ignore or be ignorant of game theory and Nash equilibria, which shows that under the right conditions purely selfish people acting rationally ought to cooperate to create outcomes that are the best achievable for everyone. (Which far from being an ivory tower theory, it describes modern capitalist society in a nutshell.)
There’s your problem. We don’t say that two things are the same if they happen to coincide under exceptional circumstances, we say they are the same if they coincide under every possible circumstance.
Ethical utilitarianism and utility based decision theory don’t coincide when someone is only a little more altruistic than a sociopath. Utilitarianism is notorious for being very demanding, so having a personal UF that coincides with the aggregate used by utilitarianism requires Ghandi level altruism., and is therefore improbable.
Likewise, decision theory can imply a CC equilibrium, but does not do so in every case.
Consequences for whom? If I violate your rights, that’s not a consequence for me. That’s one of the ways in which ethical utilitarianism separates from personal decision theory.
I don’t understand the question. “For whom” doesn’t matter. If I take an action, the world that results as a consequence has an entity who feels their rights are violated. When I sum over the utility of that world, that rights violation is a negative term, if I’m the kind of person cares about people’s rights (which I am, but is a *separate* issue).
For “the ends don’t justify the means” to mean something, it implies that there is something of intrinsic negative morality in the actions I take, even if the results are identical. I argue that this is nonsense—if there was any real, non-deontological difference you could point to, then that would be part of the utility calculation.
It matters because your ethical/decision theory will give different results depending on whose utilities you are taking into account.
If I take an action, the world that results as a consequence has an entity who feels their rights are violated. When I sum over the utility of that world, that rights violation is a negative term, if I’m the kind of person cares about people’s rights (which I am, but is a separate issue).
It’s the heart of the issue. If you don’t care about their rights, but they do, then you will violate their rights.
If there is some objective notion of the negative utility that comes from a rights violation, you will violate their
their rights unless your personal UF happens to be exactly aligned with the objective value.
For “the ends don’t justify the means” to mean something, it implies that there is something of intrinsic negative morality in the actions I take, even if the results are identical
You can’t calculate what the ultimate results are.
You have to use heuristics. That’s why there is a real paradox about the trolley problem. The local (necessarily) calculation says, that killing the fat man saves lives, the heuristic says “dont kill people”
Utilitarianism uses a version of global utility that is based on summing individual utilities.
If you could show that some notion of rights emerges from summation of individual utility, that would be a remarkable result, effectively resolving the Trolley problem.
OTOH, there is a loose sense in which rules have some kind of distributed utility, but if that not based on summation of individual utilities, you are talking about something that isn’t utilitarianism, as usually defined.
I’m not sure that was Dominic Cummings’ strategy? AFAICT Dominic Cummings is/was a regular here on LessWrong, from perhaps as far back as the OvercomingBias days, though I don’t know his account handle. Look at his blog post which quotes Eliezer Yudkowsky first, and is chalk-full of LessWrong-defined non-standard vocabulary and phrasing. He’s not (just) advertising to rationalists, that’s how he always communicates.
My understanding is that he was, at minimum, a long-time lurker here who some point he decided, pretty much on his own, that the UK leaving the EU was objectively the best outcome for “reasons”.. and then used all the tricks in the rationalist dark-arts book to make it happen. He was successful at winning, and radically changed the course of a major country.
We talk a big talk about instrumental rationality, about rationalist superpowers… but application has been a bit of a mixed bag to be honest. And yet here’s a rationalist who upturned global politics singlehandedly, and credits LessWrong with his success.
I would suggest setting aside ALL feelings about Brexit from your mind (politics is the mind-killer) and studying from an objective viewpoint what Dominic Cummings has done. There is something in the process there that ought to be emulated, even if you disagree with the instrumental outcome.
I see your point, but the outcome is important, if you want to improve things, not just become famous for changing them.
Isn’t it a commonly held belief here that the ability to achieve goals (rationality) is orthogonal from the content of those goals (morality)?
Burning down a building is easier than constructing it.
People are celebrating Dominic Cummings for changing the building. I’d like to wait until it turns out what specific kind of change it was.
In the meanwhile, I accept the argument that even burning down the building requires more skills and agency than merely talking about the building. In this way, Dominic Cummings has already risen above the level of the rationalist plebs. But how high, that still remains to be seen.
Don’t confuse the consequences of the outcome with the things that had to be achieved to accomplish it. He got a national referendum on the ballot and got it to pass, despite the establishment (political parties, media, and businesses) being uniformly against it. This is not comparable to burning down a building. Rather it was a precision campaign to identify what it would take to convince a majority of the voting population to adopt a radical political agenda.
Change “Brexit” to “basic income,” “universal single-payer healthcare,” “tuition-free higher education,” “carbon-free emissions” or whatever your preferred legislative objective would be, and the difficulty and techniques would be the same. It was a large accomplishment worth studying.
It’s not obvious to me that either the difficulty or the techniques would be the same for those other objectives as for Brexit. A canny political operative uses techniques appropriate for specific goals, after all.
If your goal is simply to get the UK out of the EU, for whatever reason and in whatever fashion, and if you don’t mind what harm you do to society in the process, then “all” you need to do is to stir up hatred and suspicion and fear around the EU and what it does and those who like it, and find some slogans that appeal without requiring much actual thought, and so forth. Standard-issue populism.
But let’s suppose you want universal basic income and you want it because you think it will help people and make society better. Then:
The “stir up fear and hatred” template doesn’t work so well, because what you’re doing isn’t a thing that can readily be seen as fighting against a shared enemy.
The “stir up fear and hatred” template may be a really bad idea even if it works, because it may do more damage to society than the reform you’re aiming for does good.
The details of what you do and how you do it may matter a lot: some versions of universal basic income might bankrupt the country, some might fail to do enough to solve the problems UBI is meant to solve, some might be politically unacceptable, etc. So you need to sell it in a way that lets a carefully designed version of UBI be what ends up happening.
The available evidence does not suggest (to me) that Cummings has a very specific version of Brexit in mind, or that he is sufficiently concerned for the welfare of the UK’s society and the individuals within it to be troubled by considerations of societal harm done by the measures he’s taken, or of whether he’s ending up with a variety of Brexit that’s net beneficial.
I would have preferred to say the foregoing without the last paragraph, which is kinda object-level political. But it’s essential to the point. When Viliam says it’s easier to burn a building down than construct it, I think he is saying something similar: if, as it seems may be the case, Cummings doesn’t actually care whether he does a lot of harm to a lot of people, then he has selected an easier task than would be faced by someone trying to bring about major reforms without harming a lot of people, and the methods he’s chosen are not necessarily ones that those who care about not harming a lot of people should emulate.
Yep. Looking around me, getting Slovakia out of EU would be relatively easier task than making it adopt UBI, for the reasons you mentioned (plus one you didn’t: availability of foreign helpers).
How clear is it that he specifically got all those things to happen? There were definitely other people involved, after all. Cummings’s own account of what happened no doubt ascribes as much agency as possible to Cummings himself, but there are possible explanations for that other than its being true.
That would imply that means are always morally neutral, which is not the case.
It’s a direct consequence of utilitarian morality.
What moral impact is there for means, other than their total consequences?
Maybe utilitarianism is wrong. If means involve rights violations, maybe they are not justified by their consequences.
You’re not applying it correctly. Rights violations are among the consequences. They’re summed as part of the utilitarian equation.
This conversation sounds like TAG uses utilitarianism to mean classic utilitarianism, where pains and pleasures are the only consequences that we care about (and rights violations are not), and like you are using it to refer to decision-theoretic utilitarianism, where the consequences can include rights violations as well.
I object that there is any real difference between the two. Classic utilitarianism is “count up all the good, subtract all the bad.” That’s exactly how I would describe decision-theoretic utilitarianism. Now the original proponents of utilitarianism also didn’t give much negative weight to rights violations, but that’s a complaint against their utility function, not utilitarianism per se.
But regardless, I think you did identify and elucidate better than I did the core disagreement here. I hope it’s resolved for TAG.
It sounds like “decision theoretic utilitarianism” was something invented here.
I think hybrid approaches to ethics have more to offer than purist approaches..and also that it is assists communication to label them as such.
Edit:
Actually , it’s worse than that. As Smiffnoy correctly states, maximising your personal utility without regard to anybody else isn’t an ethical theory at all, so it continues the confusion to label it as such.
That only describes a solipsist or sociopath’s utility function. All things being equal, I would like for you to be happy, strange person on the Internet who is reading this. Maximizing my own utility function means preferring outcomes where everyone is happy, because I value those outcomes.
Also Smiffnoy seems to ignore or be ignorant of game theory and Nash equilibria, which shows that under the right conditions purely selfish people acting rationally ought to cooperate to create outcomes that are the best achievable for everyone. (Which far from being an ivory tower theory, it describes modern capitalist society in a nutshell.)
There’s your problem. We don’t say that two things are the same if they happen to coincide under exceptional circumstances, we say they are the same if they coincide under every possible circumstance.
Ethical utilitarianism and utility based decision theory don’t coincide when someone is only a little more altruistic than a sociopath. Utilitarianism is notorious for being very demanding, so having a personal UF that coincides with the aggregate used by utilitarianism requires Ghandi level altruism., and is therefore improbable.
Likewise, decision theory can imply a CC equilibrium, but does not do so in every case.
Consequences for whom? If I violate your rights, that’s not a consequence for me. That’s one of the ways in which ethical utilitarianism separates from personal decision theory.
I don’t understand the question. “For whom” doesn’t matter. If I take an action, the world that results as a consequence has an entity who feels their rights are violated. When I sum over the utility of that world, that rights violation is a negative term, if I’m the kind of person cares about people’s rights (which I am, but is a *separate* issue).
For “the ends don’t justify the means” to mean something, it implies that there is something of intrinsic negative morality in the actions I take, even if the results are identical. I argue that this is nonsense—if there was any real, non-deontological difference you could point to, then that would be part of the utility calculation.
If I feel that I have a right to a swimming pool, does your failure to buy me a swimming pool mean that a right has been violated?
It matters because your ethical/decision theory will give different results depending on whose utilities you are taking into account.
It’s the heart of the issue. If you don’t care about their rights, but they do, then you will violate their rights.
If there is some objective notion of the negative utility that comes from a rights violation, you will violate their their rights unless your personal UF happens to be exactly aligned with the objective value.
You can’t calculate what the ultimate results are. You have to use heuristics. That’s why there is a real paradox about the trolley problem. The local (necessarily) calculation says, that killing the fat man saves lives, the heuristic says “dont kill people”
Utilitarianism uses a version of global utility that is based on summing individual utilities.
If you could show that some notion of rights emerges from summation of individual utility, that would be a remarkable result, effectively resolving the Trolley problem.
OTOH, there is a loose sense in which rules have some kind of distributed utility, but if that not based on summation of individual utilities, you are talking about something that isn’t utilitarianism, as usually defined.
Source? I’ve googled his name and LessWrong, but can’t find him saying anything about it.