A Trolley Problem Dialogue

Epistemic status: bad and wrong. Not worth reading.

Oberstein: The ethical thing to do is to pull the lever. The result is one dead person instead of five.

Kircheis: I guess. But who set up this situation? Why should I, if I have any agency at all, use it to indulge some psychopathic moral psychology experimenter's fancies?

Oberstein: Regardless of how you got into the situation, you should operate in the manner that results in the most lives saved.

Kircheis: I'm sure it's not right to treat human lives as tokens to be traded...

Oberstein: This is the same situation. You take the action that results in one corpse rather than five.

Kircheis: It's completely different! When you have to do it with your own hands, the emotional impact of that action will haunt you for the rest of your life! You know as well as anyone how much easier it is to kill someone with the push of a button.

Oberstein: Suppose the permanent emotional impact is so great that it's just as bad as dying. Then there's still the equivalent of two corpses, instead of five. Would you not sacrifice your precious feelings for the greater good?

Kircheis: You know I would.

Oberstein: Yes.

Kircheis: It's a different situation. But even if it weren't, my notion of justice doesn't allow killing people to save others. It's barbaric, or heartless. People like you can never understand it, because you think it's as simple as setting certain utilitarian weights to infinity, to the detriment of consequences. But it's not. You'll never understand deontology if you think of it as a perverse form of utilitarianism.

 At this point, Grognor materializes in a puff of manliness

Kircheis: What the hell? Who are you?

Grognor: I know it is gauche for an author to appear in his own platonic dialogue, but Kircheis is right, but he isn't smart enough to understand why. So I'm going to enhance his intelligence so he can see it. *BZZHORT* Okay Kircheis, now you can explain it to him.

Oberstein: Kircheis's position always leads to absurdities. Mine only leads to repugnances. I don't see how his can be right.

SuperKircheis: I see. Part of the reason people have different intuitions about what to do in trolley problems is that they fight the hypothetical to different degrees. People like Oberstein allow themselves to work within the assumptions of the philosopher's case, whereas people like me do not. And we are right not to, even though we usually don't know it or even realize we are doing it. Because decisions don't happen in a vacuum, and we shouldn't pretend they do.

Oberstein: Even if you can't see all the consequences of your actions, you should still act in the best interests of all within the consequences you can see, subject to your own moral and empirical uncertainty.

SuperKircheis: You are always so proud of seeing everything with those mechanical eyes. When you decided to let millions die at Westerland, did you foresee the long-term consequences of such a precedent? Did you not think of the stain it would leave on His Majesty's honor?

Oberstein:

SuperKircheis: If you say you did foresee, and sincerely believed it was worth the price, I'll believe you. But you're not seeing all the consequences of killing people in trolley problems. Your robot eyes have a myopic dysfunction.

Oberstein: Oh?

SuperKircheis: By setting a precedent where fat men are pushed off of overpasses, you make fat men afraid to cross overpasses. By being the sort of person who plays along with trolley problems, you give psychopathic moral psychology experimenters an incentive to give you trolley problems. In general, when you let yourself be exploited, you let yourself be exploited.

Grognor: So the correct action in the trolley problem where you're just flicking a switch depends on whether this implies some sort of submission to a hostile agent. Those unfortunate enough to just happen to find themselves in the position of having to choose between one death and five deaths should choose one, in situations where refusing to choose would be like refusing to allow time to pass. However, I find the epistemic state where I know exactly how many corpses will result from a decision pretty unlikely. I'm pretty confused in general.

Oberstein: I have no morals, only goals. I don't understand how I ended up in a dialogue about ethical philosophy. But, since I'm here. From an ethical point of view, don't you have a responsibility to make the best of your situations, regardless of how you got into those situations?

SuperKircheis: No, because how you respond to situations influences whether or not you get into those situations, in worlds where there are other agents who do not have the same goals as you do. I think your failure to appreciate this is why you lied to Reinhard that day. He should have punished you severely for it.

Oberstein: So you're saying evil agents who set up inevitable deaths can't simply shift culpability onto people involved in the situations they create? That is interesting. But the people involved must still choose. Refusing to choose is like refusing to allow time to pass. And they should choose based on the expected consequences.

SuperKircheis: I won't say, "To hell with the consequences!" because I'm smart enough now to know that they matter. I can't explain updateless decision theory to you right now, so let's stick with the framework where the expected consequences are the only thing that matter. If you'll allow it, the whole reason our ethical intuitions and ratiocinations take the form of strong rules instead of just figuring out what's going to happen is because we can't figure out what is going to happen. And I'm smart enough now to know that there is no amount of intelligence sufficient to figure it out. Even the very wise cannot see all ends. Well, that's half of the reason. The other half is that we constantly delude ourselves about the consequences and need strong injunctions to prevent self-serving biases from taking over our decision process. Anyway, you need an incredibly high standard of certainty before making decisions on a naive utilitarian basis is justifiable, one that mere humans, even ones as smart as I currently am, cannot attain.

Oberstein:

SuperKircheis: You're also not even taking into account structural uncertainty and metamoral uncertainty.

Oberstein: You'll have to explain what those are.

SuperKircheis: Even that wouldn't be enough, you still have to have at least a cursory understanding of the game-theoretic foundations of morality, the etiology of...

 Grognor presses the button on his device again. *BZZORRRT*

SuperOberstein: Ah. I understand now. You were right all along.

Grognor: Isn't it nice when one of these doesn't end in aporia!

My Weirdest Ethical Belief

Originally published June 25, 2016

I think male homosexuality is good and female homosexuality is bad. Mostly due to asymmetries in the sexual marketplace, the surplus of men. The upshot if you agree with me is that if you are a bisexual of either gender you should limit yourself to men. This is doubly good: it occupies a given man and frees up a theoretical woman. Additionally, putting this into practice should help control overpopulation (and be accord with antinatalism, which I also hold), though this isn't why I hold the belief at all.

Engaging the Only Reasonable Meat-Eaters

Originally published December 1, 2014

People who have even slightly non-terrible and reflective reasons for their decisions are so rare they statistically don't exist. Even so, I pay attention to a really select crowd, so among them I have discovered an entire three people whose nonvegetarianism is reasoned and principled: Robin Hanson, Katja Grace, and Paul Christiano. It's no coincidence that they are these; they are deeply influenced by each other.

Make no mistake, dear reader: I am talking about you. You do not have good reasons for eating meat. If you don't eat meat, you almost certainly have terrible reasons for that. I am not interested in engaging with terrible people and their terrible reasons. Yes, this does mean I am not interested in engaging with almost anyone.

I am socially (though not epistemologically, you'll note) obligated to explain something in the previous paragraph. I am not an advocate nor an ideologue; I just want myself and others to make the right decisions for the right reasons. I will not try to be convincing as opposed to correct. Almost all vegetarians are vegetarian for stupid reasons, like "dignity of animals", or an aversion to intentional killing. As a rule they don't even know their true reasons. Even David Fucking Chalmers has crazy reasons for his vegetarianism. Getting the right answer for the wrong reason scores zero ethics points. That's just moral luck. Of course, meat-eaters have even worse reasons. I am not interested in engaging with terrible arguments that seem good. Just actually good ones. Now then.

Actually I'm going to put yet another caveat here. You won't find in this post a single argument that eating meat is wrong. Only arguments that these people's arguments that eating meat isn't wrong are wrong. Please try to understand this.

From Hanson's Vegan Compromise:

Let us seek principles that can account for most of our acts, then try to change the other acts to conform with such easier principles.

This is explicit that his conclusions are based on weaker, easier principles than ones he could have adopted. I'd be tempted to remove Hanson from my short list of reasonable, principled nonvegetarians if not for a much stronger reason of his. Anyway one of his weak principles is "We don’t care much about most animals, even smart ones." It's wrong to use this as an unexamined principle. By default you should care about things that can suffer. This is a basic is/ought confusion.

But that's not Hanson's strongest argument. This is. Hanson believes animals live lives worth living. I strongly disbelieve this. Humans don't even usually live lives worth living. Suffering in the world outweighs joy. Arguing the previous sentence is beyond the scope of this post. Don't believe the opposite because I'm not doing so. Robert Wiblin posted a cursory examination on Robin's own blog. For more see The View From Hell and Utilitarian Essays. It is possible for reasonable people, such as Hanson, to agree with Hanson and not me here.

So much for Robin Hanson.

Katja Grace:

The real question is not whether the cost to you is small, but whether you could do more good for the same small cost.

Similarly, when deciding whether to donate $5 to a random charity, the question is whether you could do more good by donating the money to the most effective charity you know of. Going vegetarian because it relieves the animals more than it hurts you is the equivalent of donating to a random developing world charity because it relieves the suffering of an impoverished child more than foregoing $5 increases your suffering.

My imaginary vegetarian debate partner objects to this on grounds that [I, Grognor, will not be using]

[...]

However whether you can trade being vegetarian for more effective sacrifices is largely a question of whether you choose to do so.

Read the post; I can't do it justice by summarizing and cherrypicking.

I agree with the premise that a limited inconvenience budget should be spent as wisely and efficiently as possible, modulo something about the inconvenience of spending more effort on allocating inconvenience. (Although it's not clear to me that inconvenience is really so limited; a better model might be that vegetarianism is a "one-time cost" of a transition period after which things return to baseline level of ambient inconvenience.) But this is absolutely the wrong way to frame the problem. Here is an alternative framing:

Be the change you wish to see in the world.—Someone

Yes, I am employing that quote. Not every cliche is wrong. The way out of an enormous collective action problem isn't to ignore it and focus on less terrifying and less inconvenient problems. It's to stop contributing to the collective action problem. Activists love to compare meat-eating to slavery, so Hell, I'll do it too. Slavery didn't end just because technology rendered it obsolete. There were people at various points in history who, for various reasons some of them moral, consciously decided not to participate in slavery. And because there were those people, others were able to abstain. A VIRTUOUS CASCADE. There was a virtuous cascade.

No, this is not a great framing. And yes, there is a nonzero but non-negligible extent to which this is orthogonal to Katja's... point. It's still a much better framing than the one she chose, and it's easy to see classes of framings which would be better than hers. Another framing is that the extra inconvenience from vegetarianism as opposed to other things strengthens the ethical bargaining position, which Grace herself explained. Consider this:

Another way you might accidentally lose more value than you save is in spending little bits of time which are hard to measure or notice.

Miraculously, the post attracted an actually good comment, by kpier.

I get the impression that she subconsciously chose this rare framing instead of the mainstream one or a different rare one as an amazingly subtle rationalization for her newfound nonvegetarianism. It's a very well hidden in her case. So much for Katja Grace.

Paul Christiano is someone who is so many levels above me that I can't actually comprehend how much better at reasoning he is than I am. If I were a lesser man I would confuse this situation with him valuing the wrong things. Christiano (and Hanson) has spent a lot of time thinking carefully about the degree to which we are in a causal bottleneck. I get this from his blog. In particular see My Outlook and Against Moral Advocacy but I would recommend the whole blog wholeheartedly if his writing style weren't stupefyingly boring.

My unjust and misleading summary is: Paul Christiano really thinks that long-run values won't be influenced much by the social fashions we can easily influence in standard ways, e.g. by being a social justice warrior or whatever. I don't have a strong response to this. My response is to say that there's enough structural uncertainty that moral improvidence is unjustified. But if there's anyone who's given the question its due, it's Paul Christiano. So much for him.

This post was written mostly in anger. It is mostly unedited. It's not written well. I am just hopelessly shouting into the void. To get it on the record. This post will never convince anyone of anything.

By the way, I know of two people who have good and not just socially viable reasons to be vegetarian: Brian Tomasik and Robby Bensinger.