Please Don't Fight the Hypothetical

Sure. It used to infuriate me, too.

I have since then adopted the habit of responding with some variation of "I observe that you're fighting the hypothetical; I infer that's because you don't trust my motives in bringing it up. That's fair. Perhaps we can drop this subject for now, and return to it at some later time when I've earned your trust?"

People are very good at fighting hypotheticals because hypotheticals are typically an attempt to elicit a response that can be used against them. This infuriates tutors in Philosophy 100 classes, of course ...

Maybe that's so, but people need to lighten up a bit. On a discussion board such as this one, if you concede a point and later think better of it, there is nothing stopping you from explicitly revising your position. And actually learning something in the process.

I don't think I disagree. Fighting the hypo is the equivalent of changing the subject. Sometimes, that's a good idea. I certainly endorse dodging loaded questions. Even better is calling people on the loaded question, if you are brave enough and think it will help. (Unfortunately, that's usually a question of relative status, not relative rationality - but trying can still sometimes raise the sanity line).

My point was only that people sometimes get confused about what is and what isn't staying on topic.

Is there a way to tell the difference between helpful hypotheticals that illustrate confusing topics in stark terms (like Newcomb's Dilemma) and malicious ones that try and frame a political argument for rhetorical purposes (I can think of some examples, I'm sure you can as well, but let's try to avoid particulars)?

Sadly, we don't live in a world where all hypotheticals are actually neutral excercises in deductive logic. In real debates it's quite common to see people constructing hypotheticals that implicitly assume their position on some issue is the correct one. If you accept one of these hypotheticals you've already lost the argument, regardless of the actual merits of the case.

I'm not sure exactly what you have in mind, but I think in that situation the best response to be explicit about what's going on.

e.g. "I would definitely choose Option A as I think that generally speaking and all things being equal, it's better not to torture puppy dogs. But so what? What does that have to do with my original point?"

Thus, when you find yourself confronted with a hypothetical based on an incoherent ontology, corrupt definitions, or other examples of confused or dishonest thinking often the only viable response is to challenge the validity of the hypothetical itself.

I generally just ironman their flawed hypothetical and carry on. Sometimes it turns out that the flaws are easily fixed and doing so does not invalidate their original point.

For a broad enough definition of "possible", the least convenient possible world is always maximally inconvenient, no matter what your epistemology or morality or decision theory is. Any formalized deductive logic is always going to have true-but-unproveable theorems. Any priors, Occam-like or not, are always going to lead to wildly incorrect posteriors when inducting based on data from a sufficiently perverse universe. Any decision theory is always going to fail against an Omega whose behavior is "see if he's running that decision theory, then punish him accordingly".

Normally that's not the end of the world. There is probably no Omega who hates us, and so we don't have to worry about our Universe being a deceptive design or about an otherwise rational decision-making process being punished because of its rationality.

But when we start answering hypotheticals, then we are potentially talking about deceptively designed universes whose conceptual existence is a deliberate attempt to punish viewpoints not held by the hypothesizer! That doesn't mean that hypotheticals aren't useful tools or shouldn't usually be answered, but it might still be reasonable to keep the possibility of exceptions in mind.

But the same happens with things like the Trolley problem or Torture v. Specks. Many "irrational" answers to hypotheticals come out of reflexes that are perfectly rational under ordinary circumstances. If the hypothetical-proposer isn't willing to allow that the underlying reflex has its uses, the hypothetical-proposer is ignorant or untrustworthy.

It depends a bit on how well the discussion is going - if it isn't trying to reach true beliefs then that is your problem, not the appropriateness of the hypothetical.

But if the discussion is going well, then knee-jerk reactions to the hypothetical aren't helpful. The fact that a reflex answer usually works is no evidence that it handles edge cases well. And if it doesn't handle edge cases well, then it might not be a very coherent position after all.

We were explicitly told to assume our knowledge is certain. Some of the circumstances assumed long chains of improbable-seeming events between sacrificing someone and saving 5 others.

Eliezer commented on this back in "Ends Don't Justify Means (Among Humans)", attempting to reconcile consequentialism with the possibility (observed in human politics) that humans may be running on hardware incapable of consequentialism accurate enough for extreme cases:

And now the philosopher comes and presents their "thought experiment"—setting up a scenario in which, by stipulation, the only possible way to save five innocent lives is to murder one innocent person, and this murder is certain to save the five lives. "There's a train heading to run over five innocent people, who you can't possibly warn to jump out of the way, but you can push one innocent person into the path of the train, which will stop the train. These are your only options; what do you do?"

An altruistic human, who has accepted certain deontological prohibits—which seem well justified by some historical statistics on the results of reasoning in certain ways on untrustworthy hardware—may experience some mental distress, on encountering this thought experiment.

So here's a reply to that philosopher's scenario, which I have yet to hear any philosopher's victim give:

"You stipulate that the only possible way to save five innocent lives is to murder one innocent person, and this murder will definitely save the five lives, and that these facts are known to me with effective certainty. But since I am running on corrupted hardware, I can't occupy the epistemic state you want me to imagine. Therefore I reply that, in a society of Artificial Intelligences worthy of personhood and lacking any inbuilt tendency to be corrupted by power, it would be right for the AI to murder the one innocent person to save five, and moreover all its peers would agree. However, I refuse to extend this reply to myself, because the epistemic state you ask me to imagine, can only exist among other kinds of people than human beings."

Now, to me this seems like a dodge. I think the universe is sufficiently unkind that we can justly be forced to consider situations of this sort. The sort of person who goes around proposing that sort of thought experiment, might well deserve that sort of answer. But any human legal system does embody some answer to the question "How many innocent people can we put in jail to get the guilty ones?", even if the number isn't written down.

As a human, I try to abide by the deontological prohibitions that humans have made to live in peace with one another. But I don't think that our deontological prohibitions are literally inherently nonconsequentially terminally right. I endorse "the end doesn't justify the means" as a principle to guide humans running on corrupted hardware, but I wouldn't endorse it as a principle for a society of AIs that make well-calibrated estimates. (If you have one AI in a society of humans, that does bring in other considerations, like whether the humans learn from your example.)

I have to admit, though, this does seem uncomfortably like the old aphorism quod licet Jovi, non licet bovi — "what is permitted to Jupiter is not permitted to a cow."

Okay... so, you are 99 percent certain of your information. Does that change your answer?

Here's another way of putting this argument

FWIW, my answer in trolley problems is usually some variant of "Well, what I would do is probably dither ineffectually between alternatives until the onrushing train takes the choice out of my hands, but what I endorse doing is killing the smaller group." (Similarly, the truth is that faced with the canonical PD I would probably do whatever the cop tells me to do, whether I think it's the right thing to do or not. Fear is like that, sometimes. But that's really not the point of the question.)

The problem is that people don't always seem to notice that they are changing the topic. If you intentionally change the topic, that's one thing. If you unintentionally change the topic, I assert you are making a mistake.

It is not my intent to argue that changing the topic is not appropriate, because it often is the appropriate response.

For some reason, people keep thinking that Torture vs. Specks was written as an argument for utilitarianism. That makes no sense, because it's the sort of thing that makes utilitarians squirm and deontologists gloat. What it is, instead, is a demand that if you're going to call yourself a utilitarian, you'd better really mean it.

EY's actual arguments for utilitarianism are an attempt to get you to conclude that you should choose Torture over Specks, despite the fact that it feels wrong on a gut level.

I choose specks, but I found the discussion very helpful nonetheless.

Specifically, I learned that if you believe suffering is additive in any way, choosing torture is the only answer that makes sense. If you don't believe that (and I don't), then your references to "negative utility" are not as well defined as you think.

Edit: In other words, I think Torture v. Specks is just a restatement of the Repugnant Conclusion

Hank, that's one of my favorite articles.

Not sure why you are getting down-voted, but if you think "biting the bullet" is equivalent to "not fighting the hypo" then I've failed to communicate to you why fighting the hypo is bad.