1455: "Trolley Problem"

This forum is for the individual discussion thread that goes with each new comic.

Moderators: Moderators General, Prelates, Magistrates

User avatar
orthogon
Posts: 3099
Joined: Thu May 17, 2012 7:52 am UTC
Location: The Airy 1830 ellipsoid

Re: 1455: "Trolley Problem"

Postby orthogon » Tue Dec 09, 2014 9:20 am UTC

Dmytry wrote:Someone should come up with a hypothetical where it is not a brainfart to sacrifice a few to save many. There's a multitude of such cases, and they're perfectly ordinary, e.g. when you choose between workers dying of falls installing solar panels on the roofs, and people dying of cancer due to the coal power plant. It's clear that we do rationally decide on the trolley problems where the sacrificial action is not a completely idiotic brainfart (proclaimed to be effective in a far fetched hypothetical); it's just the brainfart "solution" examples where we are reluctant to sacrifice. (Ditto for the healthy patient being taken apart for organs, if you do this then people would be afraid of using healthcare. Also a brainfart solution to a problem)

Yeah, I don't think you should pick the oxymoronic "healthy patient". You should pick people at random, like jury service. That has the random-victim element of the plane crashing in the suburbs, but with the hands-on homicide of the fat man.

I don't think accept that philosophers or psychologists or whoever does these experiments are being "smug". I imagine that they like these scenarios because they themselves intuitively want to pull the lever but not push the fat man. I remember reading a passage where Kahneman or Pinker (can't remember which) explains that what normally happens is that the researchers dream up these cases and try them on themselves and each other; almost always the rigorous testing with volunteer subjects simply confirms their own reaction. We're looking for human universals, after all.
xtifr wrote:... and orthogon merely sounds undecided.

User avatar
Samik
Posts: 511
Joined: Mon May 02, 2011 11:14 am UTC

Re: 1455: "Trolley Problem"

Postby Samik » Tue Dec 09, 2014 8:26 pm UTC

Copper Bezel wrote:
Coyoty wrote:
Samik wrote:I have a question. What do people think is the purpose of thought experiments?


Manipulation. The questioner is examining the answerer's responses to find what pushes the answerer's buttons and how to manipulate the answerer. Also, the experiments themselves are a means of manipulation, for getting the answerer to change his mind from one conclusion to a different one. By proposing value judgments in different orders, the questioner can get the answerer to question his own values and be manipulated to the questioner's goal.

Also known as the Socratic Method.

Agreed (mostly*).

So, after presenting a thought experiment to a subject, and hearing their response, we then alter the parameters of the experiment, according to their response, to see if we can achieve a different response.



1.) For the thought experiment to have the greatest chance of being informative, our changes should try to address the main points of the subject's response as directly as possible.

I.E. if we present the Fat Man Trolley Problem, and they say they wouldn't push because they would be afraid of getting arrested, we don't respond, "Ok, what if there were ten people on the tracks instead of five?" We respond, "Ok, what if you somehow knew there was no chance of getting in trouble?"

The subject then gives the modified problem further consideration, eventually makes a response, and we repeat the process.


2.) An ideal subject would be willing and able to embrace any parameter, no matter how abstract or arbitrary.

I.E. it's not necessary to explain the exact mechanism by which you were certain you wouldn't get in trouble. The purpose is simply to ask, "Does your intuition remain if that possibility were removed?" Ideally, the subject is willing to play along and give the problem further consideration with the new condition(s).

While precision is important, too much detail can actually introduce noise to the response. For example, the subject who wonders how the people got tied to the track, and concludes that they bear some responsibility for being there, when the problem hasn't said anything about that, and being "tied" to the tracks was just the arbitrary mechanism by which their inability to escape was established as certain.



So, Copper Bezel says:

Copper Bezel wrote:at some point, it becomes less about what was "right" and more about what I could live with.


I then add the parameter that there is no chance of you having to "live with" your action. Perhaps you die, with certainty, in any possible scenario. Perhaps your memory is wiped. Perhaps, [insert some other arbitrary mechanism here].

Copper Bezel responds that that is "too abstract or dodges the point".

My question is: Why is that too abstract?

The parameter: [You have certainty that you will not live to be able to reflect on your decision] is no more or less abstract than [you have certainty you will not get in trouble], or [You have certainty the victims cannot escape] or [you have certainty pushing the fat man will stop the trolley] or [you have certainty pushing the fat man is the only way to stop the trolley].


*I would characterize the process less as the "questioner" trying to "manipulate" the answerer towards the "questioner's goal" (though I'm sure some do try to use thought experiments for that purpose) and more just "examination" of the answerer.
Last edited by Samik on Tue Dec 09, 2014 8:56 pm UTC, edited 2 times in total.

User avatar
Samik
Posts: 511
Joined: Mon May 02, 2011 11:14 am UTC

Re: 1455: "Trolley Problem"

Postby Samik » Tue Dec 09, 2014 8:41 pm UTC

Samik wrote:2.) An ideal subject would be willing and able to embrace any parameter, no matter how abstract or arbitrary.

I.E. it's not necessary to explain the exact mechanism by which you were certain you wouldn't get in trouble. The purpose is simply to ask, "Does your intuition remain if that possibility were removed?" Ideally, the subject is willing to play along and give the problem further consideration with the new condition(s).


Further thoughts on this:

I have noticed that subjects of thought experiments sometimes respond as if they are being "attacked". When a subject states an intuition, and the presenter then makes a focused alteration to the experiment to see if they can change the intuition, the subject will sometimes get defensive, and refuse to continue to "play along".

To me, the presenter and the subject should both be on the "same team", so to speak. They should both have an equal interest in examining the subject's intuition. The subject should not be afraid of having their mind changed, finding an inconsistency, or reaching a point where they are not immediately sure how to answer. Nor should the presenter have the intent to specifically change the subject's mind on anything in particular; the presenter should be results-agnostic, simply trying their best to describe new parameters that have a decent chance at provoking further thought from the subject.


So, in the case of the subject who's initial response is that the people "tied" to the track are "implicated" somehow, the new temporary goal of both the presenter and the subject should be to explore that intuition of "implication". The presenter should now look for ways to alter the experiment to remove any idea of "implication", or creatively alter its magnitude or character, and the subject should be willing to play along and see where it goes.

User avatar
Pfhorrest
Posts: 5474
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1455: "Trolley Problem"

Postby Pfhorrest » Tue Dec 09, 2014 11:01 pm UTC

I mostly agree with you Samik, but one important caveat to your second point is that sometimes the subject may justifiably say that they are unable to conceive of a situation where (or perhaps more strongly, believe it is impossible that) one of those parameters was true, and so cannot answer what they would do in a situation that (in their judgement) could not occur.

Say for example we were discussing some kind of religious hypothetical about what I, being the subject in this example, would do if threatened with various kinds of divine punishment in the afterlife for my behavior. I would initially respond that I'm an atheist and wouldn't take any such threats credibly. The questioner might then add the parameter "supposing you knew for certain that God existed, afterlives happened, and that God had threatened such punishment in the afterlife...". I would then respond that I don't believe it's possible to be that certain; it is simply inconceivable to me that any religion could present any kind of evidence that would make me believe in a genuine god and a genuine afterlife. Some evidence could conceivably convince me to varying degrees of confidence that there is some kind of powerful being who could somehow have access to my mind after the death of my body as I know it by some mechanism or another, and I could discuss how I would react to credible threats from such a being, but now we're getting pretty far from the connotations of "God" and "afterlife", and closer to "what if advanced aliens showed up, claimed to have created mankind, and threatened to download our brains and torture us for eternity if we didn't do what they said"; which in turn is a lot closer to "what if a guy in an alley demanded your money at gunpoint", in that the question is no longer really about religion at all and more just about how I'd react to threats.
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

User avatar
Samik
Posts: 511
Joined: Mon May 02, 2011 11:14 am UTC

Re: 1455: "Trolley Problem"

Postby Samik » Tue Dec 09, 2014 11:48 pm UTC

Pfhorrest wrote:I mostly agree with you Samik, but one important caveat to your second point is that sometimes the subject may justifiably say that they are unable to conceive of a situation where (or perhaps more strongly, believe it is impossible that) one of those parameters was true, and so cannot answer what they would do in a situation that (in their judgement) could not occur.

Granted.

Pfhorrest wrote:Say for example we were discussing some kind of religious hypothetical about what I, being the subject in this example, would do if threatened with various kinds of divine punishment in the afterlife for my behavior. I would initially respond that I'm an atheist and wouldn't take any such threats credibly. The questioner might then add the parameter "supposing you knew for certain that God existed, afterlives happened, and that God had threatened such punishment in the afterlife...". I would then respond that I don't believe it's possible to be that certain; it is simply inconceivable to me that any religion could present any kind of evidence that would make me believe in a genuine god and a genuine afterlife. Some evidence could conceivably convince me to varying degrees of confidence that there is some kind of powerful being who could somehow have access to my mind after the death of my body as I know it by some mechanism or another, and I could discuss how I would react to credible threats from such a being, but now we're getting pretty far from the connotations of "God" and "afterlife", and closer to "what if advanced aliens showed up, claimed to have created mankind, and threatened to download our brains and torture us for eternity if we didn't do what they said"; which in turn is a lot closer to "what if a guy in an alley demanded your money at gunpoint", in that the question is no longer really about religion at all and more just about how I'd react to threats.

I understand the argument. If I put such a question to you, and you responded in that way, I would not feel that you were just "refusing to play along" or such.


But, to me, being the natural devil's advocate that I am, even these kinds of questions are perfectly acceptable launching points for all kinds of philosophical investigations. Again, what's important is that neither the presenter nor the subject have a hard and fast agenda. If the presenter is trying to get the subject to conclude [that one should behave a certain way in the real world] "under threat of divine punishment", or something similar, based on [these given spiritual conclusions], then sure, you're prefectly justified in rejecting the conclusions the presenter is asking you to accept as givens. (More generally, I would say you're perfectly justified in resisting any line of questioning where the presenter has an obvious agenda - back to the idea of an ideal encounter having all participants "on the same team".) ((More realistically, the stronger the presenter is pushing a certain agenda, the more likely I am to take the bait and try to push back, whether I agree with them or not. :twisted: ))

But if you're just two nerds who want to have an interesting debate, well, why not pose the question and then begin fiddling with the parameters? Perhaps in trying to nail down the parameters sufficiently for the subject to find the question answerable, the presenter gives a characterization of "god" that is indistinguishable from "advanced aliens [who] showed up, claimed to have created mankind, and threatened to download our brains and torture us for eternity if we didn't do what they said". Now the subject has to try to figure out how the threat of that outcome affects their feelings on whether or not to push the fat man, and the subject and presenter both have to try to figure out if that result from the presenter's efforts is meaningful.

Or, perhaps, in trying to nail down the parameters sufficiently for the subject to find the question answerable, the presenter accidentally performs a reductio ad absurdum, and realizes that they can't actually fully explain what it was they were trying to describe in the first place.



Edit: Anyway, I realize that basically all of this post is one big tangent that has little to nothing to do with the content of the rest of this thread. Sometimes (often) I just can't not ramble. I know that. I'm working on it.

User avatar
Coyoty
Posts: 195
Joined: Wed Jun 06, 2012 5:56 pm UTC

Re: 1455: "Trolley Problem"

Postby Coyoty » Wed Dec 10, 2014 8:37 pm UTC

orthogon wrote:I don't think accept that philosophers or psychologists or whoever does these experiments are being "smug". I imagine that they like these scenarios because they themselves intuitively want to pull the lever but not push the fat man. I remember reading a passage where Kahneman or Pinker (can't remember which) explains that what normally happens is that the researchers dream up these cases and try them on themselves and each other; almost always the rigorous testing with volunteer subjects simply confirms their own reaction. We're looking for human universals, after all.


I read a science fiction story where the writer was visited by a time traveler who was a fan of his work. The traveler posed a thought experiment and asked the writer whether it would be a moral thing to do. The traveler would visit occasionally with different thought experiments. The writer concluded finally that they were actual experiments the traveler had done, and was looking for validation or forgiveness from the writer to ease his conscience.

User avatar
Quizatzhaderac
Posts: 1821
Joined: Sun Oct 19, 2008 5:28 pm UTC
Location: Space Florida

Re: 1455: "Trolley Problem"

Postby Quizatzhaderac » Wed Dec 10, 2014 11:13 pm UTC

PinkShinyRose wrote:We could take this further, but we could just as well ask soldiers ordered to murder innocent civilians on the ground and soldiers to do the same from their home base using a drone and compare how they report they felt while doing it (and maybe compare it with a pilot bombing civilian targets). Isn't there plenty of real world data on this? Without the saving people part, that is.
Drone pilots experience post traumatic stress at similar rates as ground troops, but I think you misunderstand how drones work. "Drones" are entirely manual, but controlled remotely. The drones have excellent cameras and strikes are proceeded by surveillance; the pilot is very aware they're targeting a human and leaving a corpse.
PinkShinyRose wrote:What do you fill in when there is no option you identify with more than any other? What are the options anyway? If it's just skin tone from 1-10 (1 being snow and 10 being soot tint or the other way around), there is always a reasonable answer, but if it asks for geographic origin type names it could become difficult if you're 50-50 and didn't grow up in an environment with people of 1 geographic origin. You could fill two options, but that often results in an invalidated answer. A fix could be an "other" option.
There are essentially six questions on race and ethnicity (framed as check-boxes) any combination of answers is permissible:
Are you American Indian/ native Alaskan?
Are you Asian?
Are you black/African-American?
Are you Hispanic?
Are you Hawaiian Native or Pacific islander?
Do you belong to another race?

So you would check Asian, black and white.

Azule was likely referring to older censuses where the answers were mutually exclusive.

As for how it's used, I'll answer in the context of home mortgage lending since I know that in detail.
When applying the borrower answers however they like, or refuses (refusal is also recorded).
If the borrower refuses the agent of the bank is required to fill out the race/ethnicity based on appearance, speech, or surname. This is done so that if there's enough information to be racist, they need to fill out this information.
The bank is explicitly forbidden to use this information but must store and preserve it.
The bank can be sued for explicitly discriminatory policies, unintentionally discriminatory policies without sound justification, discrimination by it's agents outside of policy.
Because banks don't like being sued their legal teams search their policies for any hint of racism. Management also polices agents of the bank to ensure they don't discriminate of their own volition.
Last edited by Quizatzhaderac on Wed Apr 20, 2016 10:08 pm UTC, edited 1 time in total.
The thing about recursion problems is that they tend to contain other recursion problems.

User avatar
PinkShinyRose
Posts: 835
Joined: Mon Nov 05, 2012 6:54 pm UTC
Location: the Netherlands

Re: 1455: "Trolley Problem"

Postby PinkShinyRose » Fri Dec 12, 2014 1:47 am UTC

Quizatzhaderac wrote:
PinkShinyRose wrote:We could take this further, but we could just as well ask soldiers ordered to murder innocent civilians on the ground and soldiers to do the same from their home base using a drone and compare how they report they felt while doing it (and maybe compare it with a pilot bombing civilian targets). Isn't there plenty of real world data on this? Without the saving people part, that is.
Drone pilots experience post traumatic stress at similar rates as ground troops, but I think you misunderstand how drones work. "Drones" are entirely manual, but controlled remotely. The drones have excellent cameras and strikes are proceeded by surveillance; the pilot is very aware they're targeting a human and leaving a corpse.

I was mostly referring to an earlier mention of physical proximity being relevant in the fat man scenario. While there is a videoconferencing-like presence with the drones, the drone pilot is not physically present and does not risk having to step over the fresh corpse to enter some building or something similar. In that sense they seem more similar to helicopter pilots than to ground troops (except that the helicopter pilots are physically on scene). Of course, we would like to compare with soldiers using melee weapons, but those were already ineffective in WWII, so they're going to be difficult to find.
Quizatzhaderac wrote:
PinkShinyRose wrote:What do you fill in when there is no option you identify with more than any other? What are the options anyway? If it's just skin tone from 1-10 (1 being snow and 10 being soot tint or the other way around), there is always a reasonable answer, but if it asks for geographic origin type names it could become difficult if you're 50-50 and didn't grow up in an environment with people of 1 geographic origin. You could fill two options, but that often results in an invalidated answer. A fix could be an "other" option.
There are essentially six questions on race and ethnicity (framed as check-boxes) any combination of answers is permissible:
Are you American Indian/ native Alaskan?
Are you Asian?
Are you black/African-American?
Are you Hispanic?
Are you Hawaiian Native or Pacific islander?
Do you belong to another race?

So you would check Asian, black and white.

Azule was likely referring to older censuses where the answers were mutually exclusive.

As for how it's used, I'll answer in the context of home mortgage lending since I know that in detail.
When applying the borrower answers however they like, or refuses (refusal is also recorded).
If the borrower refuses the agent of the bank is required to fill out the race/ethnicity based on appearance, speech, or surname. This is done that if there's enough information to be racist they need to fill out this information.
The bank is explicitly forbidden to use this information but must store and preserve it.
The bank can be sued explicitly discriminatory policies, unintentionally discriminatory policies without sound justification, discrimination by it's agents outside of policy.
Because banks don't like being sued their legal teams search their policies for any hint of racism. Management also polices agents of the bank to ensure they don't discriminate of their own volition.

This seems like a sensible system, the older censuses you refer to do not but that's probably why they were changed. (Although Asian is vague: is it specifically East-Asian? Does it include Anatolian Turkish people? Does it include Jews (although now that I think about it, the Bible/Torah seems to say they are actually from Egypt and not from Israel which would make them African)?)

User avatar
Pfhorrest
Posts: 5474
Joined: Fri Oct 30, 2009 6:11 am UTC
Contact:

Re: 1455: "Trolley Problem"

Postby Pfhorrest » Fri Dec 12, 2014 2:13 am UTC

PinkShinyRose wrote:This seems like a sensible system, the older censuses you refer to do not but that's probably why they were changed. (Although Asian is vague: is it specifically East-Asian? Does it include Anatolian Turkish people? Does it include Jews (although now that I think about it, the Bible/Torah seems to say they are actually from Egypt and not from Israel which would make them African)?)

By the racial reckoning of the time, ethnic Jews are of the "Semite" race, as distinguished from the "Hamite" and "Japethite" races, which three would translate into more recent terminology as something like "Asian", "African", and "European", respectively. So by their own understanding of it (or at least, by their ancestors' understanding of it), Jews are "Asian".
Forrest Cameranesi, Geek of All Trades
"I am Sam. Sam I am. I do not like trolls, flames, or spam."
The Codex Quaerendae (my philosophy) - The Chronicles of Quelouva (my fiction)

User avatar
addams
Posts: 10325
Joined: Sun Sep 12, 2010 4:44 am UTC
Location: Oregon Coast: 97444

Re: 1455: "Trolley Problem"

Postby addams » Fri Dec 12, 2014 2:33 am UTC

Pfhorrest wrote:
PinkShinyRose wrote:This seems like a sensible system, the older censuses you refer to do not but that's probably why they were changed. (Although Asian is vague: is it specifically East-Asian? Does it include Anatolian Turkish people? Does it include Jews (although now that I think about it, the Bible/Torah seems to say they are actually from Egypt and not from Israel which would make them African)?)

By the racial reckoning of the time, ethnic Jews are of the "Semite" race, as distinguished from the "Hamite" and "Japethite" races, which three would translate into more recent terminology as something like "Asian", "African", and "European", respectively. So by their own understanding of it (or at least, by their ancestors' understanding of it), Jews are "Asian".

Yeah.
That's so funny.

Has anyone checked with the Asians?
Do they like being Lumped together with everyone that does not Look ethnicity African or North West European?

I guess it depends upon what Asian we ask.
The Eastern Indian people from England seem to think they are Asian.

Do they like being lumped in with Japanese people?
Are the Japanese going to believe the East Indians and the Jews are Asian, just like they are.

Maybe..
Big Ole' Maybe..

The Japanese people will easily believe East Indians are Asian, like Thai People and Vietnamese and Mong.
Not Asian like Japanese. No one is Asian like Japanese, except Chinese.

Maybe that's Why we hear of Japan/China discord.
They each think they know how to be Asian the Right and Proper way.
Anything else is Foreign.

Now they have to attempt to understand How Jews are Asian.

Explain it to me;
I'll Translate!

It'll be Great!
I'll understand it, finely.

And; We will get the Japanese all straightened out.
The Japanese can explain it to the Chinese.
Life is, just, an exchange of electrons; It is up to us to give it meaning.

We are all in The Gutter.
Some of us see The Gutter.
Some of us see The Stars.
by mr. Oscar Wilde.

Those that want to Know; Know.
Those that do not Know; Don't tell them.
They do terrible things to people that Tell Them.

User avatar
addams
Posts: 10325
Joined: Sun Sep 12, 2010 4:44 am UTC
Location: Oregon Coast: 97444

Re: 1455: "Trolley Problem"

Postby addams » Fri Dec 12, 2014 2:46 am UTC

Quizatzhaderac wrote:
PinkShinyRose wrote:We could take this further, but we could just as well ask soldiers ordered to murder innocent civilians on the ground and soldiers to do the same from their home base using a drone and compare how they report they felt while doing it (and maybe compare it with a pilot bombing civilian targets). Isn't there plenty of real world data on this? Without the saving people part, that is.
Drone pilots experience post traumatic stress at similar rates as ground troops, but I think you misunderstand how drones work. "Drones" are entirely manual, but controlled remotely. The drones have excellent cameras and strikes are proceeded by surveillance; the pilot is very aware they're targeting a human and leaving a corpse.

oh, You are having a serious conversation.
Yes. Some of the pilots of unmanned, remote controlled weapons feel the weight of what the machine did while under their control.

I have spoken to such a person.
I felt so helpless.

He wanted changes in Policy.
I could not really understand what he was talking about.

Neither of us could make a fuss.
He was Not supposed to be telling me.

That was long ago, before Drones were Common Knowledge.
Not every man that Kills at a distance, through a screen feels bad about it.

Some have been in training as part of their lives for most of their lives.
There may be a very good reason little American Boys and Girls play such Violent Games.
Life is, just, an exchange of electrons; It is up to us to give it meaning.

We are all in The Gutter.
Some of us see The Gutter.
Some of us see The Stars.
by mr. Oscar Wilde.

Those that want to Know; Know.
Those that do not Know; Don't tell them.
They do terrible things to people that Tell Them.

User avatar
Quizatzhaderac
Posts: 1821
Joined: Sun Oct 19, 2008 5:28 pm UTC
Location: Space Florida

Re: 1455: "Trolley Problem"

Postby Quizatzhaderac » Fri Dec 12, 2014 5:15 pm UTC

PinkShinyRose wrote: (Although Asian is vague: is it specifically East-Asian? Does it include Anatolian Turkish people? Does it include Jews (although now that I think about it, the Bible/Torah seems to say they are actually from Egypt and not from Israel which would make them African)?)
These categories picked at a time when America had much smaller populations from the middle east and Indian sub-continent. Indians typically check Asian. Arabs check white or other; apparently they're considering adding an arab/ north African category. American Jews are predominately of European decent and are typically white for census purposes (all thought this does fail to track a category of discrimination ).
The thing about recursion problems is that they tend to contain other recursion problems.

User avatar
PM 2Ring
Posts: 3715
Joined: Mon Jan 26, 2009 3:19 pm UTC
Location: Sydney, Australia

Re: 1455: "Trolley Problem"

Postby PM 2Ring » Mon Apr 22, 2019 3:27 pm UTC

Image


Return to “Individual XKCD Comic Threads”

Who is online

Users browsing this forum: No registered users and 108 guests