top | item 31996235

Absurd Trolley Problems

1296 points| sebg | 3 years ago |neal.fun

920 comments

order
[+] jgerrish|3 years ago|reply
This kind of highlights a subtle interactive design or user experience component of the Trolley problem.

Part of the Trolley Problem is that the choice is between an action or inaction. But these problems, or the first one at least, have two buttons.

So you're making a choice. Yeah, some might argue it's a technicality. But what if you put the Do Nothing choice on a timer?

What if you made it an asynchronous "Problem of the Day". No action by the end of the day triggers Do Nothing, etc.

Lots of cool, interesting design choices. There's an unstated subtle ultimatum hidde. in these problems, but still cool.

Sorry, geeking out.

[+] rax0m|3 years ago|reply
I did the problems under the assumption that "pulling the lever" was an act, while doing nothing was not acting. Implied legal (and moral) liabilities made a difference in my choices.
[+] makoto12|3 years ago|reply
I think this is a really good point. Definitely makes inaction feel like a choice, which is fundamental to the problem
[+] llimos|3 years ago|reply
This is exacerbated by the summary at the end, giving you your "kill count".

For cases where you did nothing, you did not kill anyone. The trolley did.

[+] layer8|3 years ago|reply
The problem is to pick a timeout long enough to allow the player to at least read through the description and understand the situation, but not so long that players gets impatient and pick the “action” because they don’t want to wait (and there are no real stakes on the website). A timeout could distort the results more than the current version.
[+] somedude895|3 years ago|reply
Would be great if you had to actually pull a lever and hold it, being able to release it at any time. The cartoon figures' faces change depending on whether they're currently in danger or not. Then see how many people change their mind mid pull like with the best friend or the cat/lobster one.
[+] stared|3 years ago|reply
This action/inaction is done nicely in Trolley Problem Inc (https://store.steampowered.com/app/1582680/Trolley_Problem_I...).

As long as the trolley moves, you can switch back.

While I have mixed opinions on the game (literally mixed, i.e. both good and bad, not a euphemism for bad), it provides more uneasy emotion than any other trolley problem game I played.

The "bad" part is that it is a paid game, working only on Windows, while at the quality of a Flash game from the 2000s from Newgrounds. The good part is that it is at the creativity level of a Flash game from the 2000s from Newgrounds.

[+] phkahler|3 years ago|reply
"If you choose not to decide you still have made a choice." - Rush
[+] alliao|3 years ago|reply
I've played some Japanese games where indecisiveness is represented as inaction and is one of the option, first time realising it definitely upped the ante
[+] Folcon|3 years ago|reply
Ok, this is really interesting, is there some UX where inaction itself stops being the norm?

For example 2 buttons, with a timer to the side, which when it hits 0, rolls a dice and picks a random outcome.

Does the above now make someone want to choose?

If not is there any mechanism that does not forcibly require a decision (ie just stating you must make a choice or place the chooser in a situation where a choice is a functional requirement, such as you're in a locked room and cannot leave until you press a button).

[+] andix|3 years ago|reply
Yeah, that one bugged me too. The trolley should move right from the beginning and go straight if you don’t pull quickly enough.
[+] Silica6149|3 years ago|reply
Another point is that a lot of these scenarios are easy for us to think through, but to be in the actual situation and physically push someone/something and potentially see someone die would make a lot more of us indecisive.
[+] mmcnl|3 years ago|reply
Also a part of the trolley problem is that your action space is limited to two actions. In practice that's never the case.
[+] daniel-cussen|3 years ago|reply
And like why is the trolley about to run over five people? And why can't you personally jump in the way? Who designed the train track? Same person that wants you to scapegoat some fatty?
[+] paxys|3 years ago|reply
Level 20 was weird. It was a choice between letting a trolley run as normal ("emits CO2, kills 3 people in accidents over 30 years") or running it into a brick wall and decommissioning it. For some reason more people picked the latter. So, they just dislike public transit? What about the emissions and death rate when everyone switches to cars instead?

Goes to show how easily the context and exact wording of a question can sway people's opinions.

[+] JonathanFly|3 years ago|reply
>Oh no! Due to a construction error, a trolley is stuck in an eternal loop. If you pull the lever the trolley will explode, and if you don't the trolley and it's passengers will go in circles for eternity. What do you do?

50% of people pull the lever?!?

I hereby declare that if I'm ever going to be stuck in a trolley for my entire life, I do NOT want the lever pulled. Toss me a smartphone charger and my life wouldn't even be that different, day to day.

[+] _nalply|3 years ago|reply
This was quite easy for me. The principles:

1. Only pull the lever if you are sure. With an action you assume responsibility for the outcome.

2. Don't believe everything what you see or are told. If you doubt do nothing. Imagine you killed many people because you have been lied to and pulled the lever. In this case it is better you did nothing.

3. An action of yours might kill someone. This is a very heavy responsibility you assumed. You need to be very sure and the odds need to be extreme. None of the problems managed to tip the odds. So whenever I was afraid that pulling the lever kills someone, I didn't pull the levers.

4. If my life is at stake, I pull the lever to save myself.

5. If I am extremely sure that no lives are at stake and I am relatively confident that pulling the lever will avoid lots and lots of damage, then I pull the lever. If I am unsure, I do nothing.

6. In all other cases responsibility is too high and I won't do anything. The idea: don't touch the damn thing.

The website told me that I decided differently than the majority of people and finally that I "have solved philosophy"

EDIT: The top comment made a good point. I had to click a button to do nothing. This made me realize that I a had a silent assumption:

7. Time to decide is short. The trolley is already coasting. So I didn't think long and hard because I didn't have time and when I was unsure I would have let the trolley coast past without having pulled the lever.

[+] mcv|3 years ago|reply
I noticed I kept switching between several principles. Possibly triggered by the details of the problem. Some of the principles:

1. If you're able to act, you are already responsible for the outcome. Inaction is a choice too. There's no intrinsic moral difference between pulling the lever or not.

2. Acting from a position of ignorance is irresponsible. Better not to act than to take the risk of making the situation worse. (This contradicts #1.)

3. The person who tied these people to the track is the one who really carries the responsibility to this tragedy, not me.

All three of these are valid, and yet contradictory to some extent.

Apparently I solved philosophy at the cost of 59 lives. Not sure it was worth that sacrifice.

[+] littlestymaar|3 years ago|reply
> 4. If my life is at stake, I pull the lever to save myself.

I find this answer really interesting, because in the entire corpus, this was the most morally unambiguous question of all, and the answer is diametrically opposed to yours.

Would have I the courage to do so in real life, I don't know, and it's probable that I would not. But from a moral perspective it's damn clear: the only people I am unambiguously allowed to kill to save someone else is myself. Killing someone to save someone else is a moral dilemma as it makes me take someone's life, but killing myself isn't.

[+] galfarragem|3 years ago|reply
The ultimate trolley problem is one where you pull the lever to save a larger amount of people that have chosen a unsafe area while punishing a smaller amount of people that have chosen a safe area (unless someone pulls the lever)... Ultimately this is politics.

Morality is not math, but, as always, things might be more foggy: lack of free choice and a really skewed ratio (e.g. punish one to save 100 million) may test the limits of morality.

[+] SilverBirch|3 years ago|reply
Yeah one of the issues with the trolley problem is how literally you take the problem. If you take it perfectly literally and say "5 vs 1" (for the pure example) no other context I think you can very easily come to a conclusion that is the opposite of what you'd do in real life, where either due to lack of certainty, lack of perfect knowledge or some iterative game theory interpretation, you would act differently.
[+] lumenwrites|3 years ago|reply
> Only pull the lever if you are sure. With an action you assume responsibility for the outcome.

Letting people die just because I don't feel super comfortable assuming the responsibility sounds pretty close to the definition of evil to me.

[+] kortex|3 years ago|reply
I found myself following a heuristic that is basically Asimov's laws:

- whenever possible, do no harm

- do not let harm occur due to inaction

- when given a choice, preserve the most amount of healthy lifespan in aggregate

- higher lifeforms are more valuable than lower ones (cat vs lobsters)

- deferred consequence is better than immediate (since it opens the door to other later interventions)

It kind of really brings to bear how much of a thematic device the 3 laws are. There's no way to make them congruent with actual, messy, real-world situations. Also why the whole "self-driving car trolly problem" is a non-issue - there will never be a situation where the "AI" has nice neat consequences and a binary choice laid out in front of it. It's always going to be some collection of "preserve life as best as possible" heuristics.

[+] ConstantVigil|3 years ago|reply
Kill count: 84

Apparently I solved philosophy? Im gonna assume that title stays the same for everyone lol.

So, explanation for why so high.

Simply put; I don't believe 5 people really got tied to those rails each time. They are in on it somehow. If so, that is sick and twisted they would tie someone else to the rails purposely. So they deserve the trolley instead.

Now on the chance that they really are all innocent; I still have another problem with it.

How did they not overpower the person tying them up? The single person I can understand. Heck 2 people even. 3? 4? 5!?

no.

Something is up. That trolley is going over the 5.

(edit: Like seriously, these people would all have to have been knocked out with some drugged food or something at a party first... And that's assuming they stay asleep, don't struggle, etc...)

[+] everfree|3 years ago|reply
I'm surprised to see the popular answer to Question 3.

> Oh no! A trolley is heading towards 5 people. You can pull the lever to divert it to the other track, but then your life savings will be destroyed. What do you do?

Over 70% chose to pull the lever and destroy their life savings.

People die of preventable causes in developing countries today. By choosing not to donate your life savings today to help them, you are choosing not to pull the Question 3 lever.

According to Givewell, it takes $4500 to save a life in Guinea. So for every $4500 of your savings that you choose not to donate to Guinea, that's one person you are choosing not to pull the lever to save. Have $45,000 in savings? That's 10 people you're choosing not to pull the lever to save.

I doubt that over 70% of respondents are regularly donating anywhere close to their life savings.

[+] nocturnial|3 years ago|reply
If you factor in the legal implications, the trolley problem becomes trivial. Do nothing. I'm not qualified nor allowed to operate train infrastructure and the legal consequences will become worse if someone dies because of something I did.
[+] lettergram|3 years ago|reply
I always found the trolley problem interesting. It usually boils down to who believes in karma or a creator or not.

Most religions have the idea that it’s different to take a life than to stand by and do nothing. For instance, you should always try to help others (save a life), pretty much above all else. However, to take a life, requires taking action. Ie standing by as someone drowns is not the same as sticking someone under water. For murder you’re damned to hell. For standing by you’ll need to repent, but it’s a lesser sin.

The trolly problem IMO is a framing problem. (1) it assumes you know the future and (2) it assumes your will is above others.

The example I typically gave people when discussing this problem is actually in this fun exercise. Imagine if the 5 people strapped themselves to the tracks. Imagine they knew they would murder and wanted to die rather than murder. When you redirect the train, you actually cause more deaths, because you didn’t know the intent of those people.

The exercise helps you decide what you value. For me I never apply my outside influence to the system, except when I can save a life without costing a life. I believe life is more valuable than pretty much anything I saw in the game.

I’m a realist and objectivist. So for me the question is “what could I live with?” And “what information do I have to make a decision?”

In reality, everyone in this situation made their bed (so to speak). So I am unwilling to ever impose my will into the system baring saving a life (without costing one).

[+] beaker52|3 years ago|reply
This might be something for my own trip, but the "trolley continuing for eternity" question was interesting to me.

I experienced it as if the universe (the creator) was asking itself (me) whether it should carry on doing itself (literally, and figuratively) or stop and bring the trolley ride of life to a complete end.

Enough philosophy for today.

returns to pretending that being a jelly-covered skeleton that has a hole in its face to put food which eventually comes out of another hole 3-inches from their magical life-creating sex organs (that they rub together to make new ones) and waking up every day to a world where they sit in front of a box that requires they press buttons in the right sequence to make their own life continue, is completely bloody normal

Happy Existential Wednesday.

[+] ravi-delia|3 years ago|reply
Ending with a contextless kill count was the perfect touch. This feels almost as profound as it is silly, but I'm not sure how meta-level the profundity is. Lovely site for sure
[+] duskwuff|3 years ago|reply
I just wish they'd included the complete kill count, not just the people. "You killed 85 people, one painting, and five lobsters." Or whatever.
[+] MrPatan|3 years ago|reply
Trolley problems suck.

Anybody's response is meaningless unless we take into account the context they have in their heads.

A very common piece of context is: "If you touch something and people die, it's your fault. If you don't touch anything and people die, you can very convincingly claim in front of a judge 'It wasn't me, I didn't touch anything!'"

That's a very different setting than "I'm a philosopher who wants to show how much I care or don't about total utility to other philosophers"

[+] adrianmsmith|3 years ago|reply
This is exactly how I've come to think about them as well. If you get involved with something, someone might blame you.

I think that's why (bad) managers often just don't seem to make any decisions. You need to know if the feature should be like X or Y, why won't they just make a decision and tell you? It's because they instinctively understand if they can get away with not making a decision (e.g. you stop waiting and just go for one of the option, or you contact someone else to make a decision) then there's less chance they'll be blamed for stuff.

[+] mikkergp|3 years ago|reply
I think the emotional context / trauma is interesting as well when considering many of the scenarios. Regardless of some sort of external validation, how will you feel after incident? The "If you touch something and people die, it's your fault" description is very clean and logical. or something like "I don't want to play god, when there's so much uncertainty", and lots of people leave it at that. But what would your conscious hoist upon you in the days and years after you made your decision. What could you _actually_ live with, not what do you think you could live with.
[+] JasonFruit|3 years ago|reply
I think trolley problems are mostly linguistic. Take this wording: "Oh no! A trolley is careening into a crowd! Should I try to turn it so it goes through the sparsest part of the crowd?" I think most people — by a wide margin — would say yes, because it's worded as still hitting the same thing, just hitting it in a way that does less damage, and we feel differently about that than hitting a different thing, "this guy" instead of "those guys".
[+] DoughnutHole|3 years ago|reply
That's not really just a difference of language, it's a different situation from the base trolley problem. You're describing a realistic situation where the potential outcomes are fuzzy and you're just trying to minimise the probability of damage. The base problem is binary - definitely let 5 people die or definitely deliberately kill 1 person.

If you believe the probabilistic and discrete versions of the problem are the same that just means you fall into the utilitarian camp when it comes to this thought experiment, and believe that the outcome is the only thing that matters.

[+] Helmut10001|3 years ago|reply
From an ethical point of view, the biggest problem for me was when there were two bad choices to be weighted against each other, e.g. 1 person on a track dies against 5 persons on the other track. In most of these cases, I "did nothing" - because I don't want to play god and I knew where this leads: There are ethical edge cases that I definitly do not want to think about. I am not super convinced about this in hindsight: E.g. Triages exist in hospitals to help doctors find the least worse outcome among many bad options. Most, this involves metrics such as number of dead patients, or number of lived years lost (e.g. prefer to safe children before elderly).
[+] langitbiru|3 years ago|reply
Well, in China they ask this trolley problem.

If you were a son, would you save your mother or your gf?

The "correct answer" is your mother.

"China's ministry of justice later posted the "correct" answer: exam writers are duty-bound to save their mothers. It would be a "crime of non-action" to choose romantic love over filial duty. "

https://www.bbc.com/news/blogs-china-blog-34377611

[+] Ekaros|3 years ago|reply
From Western perspective that might seem strange.

But I think this trolley test might be extended to pull a lever to kill man instead of a woman. Which to I think we have some very sad notions of equality...

[+] breuleux|3 years ago|reply
As a compatibilist, I don't know what to do with level 28 ("Oh no! A trolley problem is playing out before you. Do you actually have a choice in this situation? Or has everything been predetermined since the universe began"). Both statements can be true! Although I guess that means there is no choice :(
[+] pwpw|3 years ago|reply
I thought it could have expanded more on animal sentience. Maybe someone can come up with better examples, but my examples would be:

- A trolley is heading towards 20 puppies. Do you pull the lever to divert it towards one person?

- A trolley is heading towards 20 pigs. Do you pull the lever to divert it towards one person?

- A trolley is heading towards one person. Do you pull the lever to divert it towards your healthy pet dog that you consider as a family member?

I think the responses would be insightful.

[+] TrapLord_Rhodo|3 years ago|reply
> Oh no! A trolley is heading towards 5 people. The lever just speeds up the trolley, which might make it less painful. What do you do?

I voted no, cause really? who the hell would SPEED UP a train to hit someeone because it 'might' make it less painful. Disregarding the million other mights that could be assumed. The logic here is astounding & yet.

33% of people agree with you, 67% disagree (20,485 votes)

I guess the lesson here is people act on assumptions as tho they are fact.

[+] bmo-at|3 years ago|reply
I was surprised about that one as well. Here is another assumption: Speeding up the train might reduce their chance of being saved by a third party. So by speeding up the trolley you might make it less painful, but also might rob them of the chance of being saved. I know this is just a thought experiment, but there are a million variables to a scenario like this and just adjusting for the first one you see is a slippery slope.
[+] tgv|3 years ago|reply
It's going to happen anyway, so there's a net positive effect of speeding it up. I can't imagine people not choosing to do so.

> I guess the lesson here is people act on assumptions as tho they are fact.

As I wrote above: these problems are too artificial to draw real conclusions. They're thought experiments. The only information you can use is given. For the rest, you have to fall back to "defaults" (each life is worth the same, etc.).

If you look at it like that, it's interesting to see how many people would opt to run over the rich man.

[+] McBeige|3 years ago|reply
I voted to speed it up. "I don't know everything so I remain cautious" is a mentality I applied for some problems, and not for others. In this case I thought the problem would be trivial/boring if considered from that mentality, so I instead made a choice between condemning someone to a 10 second life of 10/10 pain and condemning someone to a 5 second life of 5/10 pain.
[+] FranticFervor|3 years ago|reply
Also voted no, felt like pulling the lever would mean I took part in the inevitable killing, even if the outcome was the same. It would also remove precious time needed for the victims to recount their lives and come to terms with their situation before dying.