> The worst case was someone's arm being badly injured, maybe permanently. On the other side was delaying the whole line for an hour, perhaps losing ~4 years of waking life in total (~25k people 1hr / 16 / 365 = 4.3y).*
I am not convinced by these kinds of calculations, and I think they generally do rationalists & their like a big disfavor.
A year of someone's life is approximately 9,000 hours.
If you spend an hour commuting to work and back for 30+ years, that's approximately the same amount of time.
But would anyone accept a magic genie offer where they agree to die a year earlier, in exchange for a magic ring that lets them teleport to work and back, skipping the commute? (Hand-waving away the various abuses of that sort of power; the genie is crafty, and will absolutely not allow you to metagame this sort of thing.)
I think not. Spending an hour waiting for, or on, a train is not the equivalent of losing an hour of your life.
I think I would actually. The tradeoff between commute time and housing location is one of the largest time/money/lifestyle decisions you constantly make over your professional life. If I expected my commute would be 30m+ every day by car, I believe I would take that trade in a heartbeat. I could completely decouple where I live from where I work.
I think that is the forbidden metagaming the parent comment is talking about. You'd still love and work in the same places and take commute time into account when choosing a place. You just wouldn't experience the commute.
> But would anyone accept a magic genie offer where they agree to die a year earlier, in exchange for a magic ring that lets them teleport to work and back, skipping the commute? (Hand-waving away the various abuses of that sort of power; the genie is crafty, and will absolutely not allow you to metagame this sort of thing.)
This is so obviously a beneficial trade that I feel like I must be missing something.
EDIT: even if you added the stipulation that I must continue to live where I currently do, and can't take advantage of this teleportation to live in a more-favourable area that is further from my work.
And what of the similar delay incurred when someone's arm does get ripped off? It's not like the train is just going to keep going on with it's regularly scheduled day. Paramedics will be called, the train will be stopped.
Besides that, this happened 11 years ago. Probably everyone involved has collectively wasted that 4.3 years just by not optimizing their time to within 1 second per day (11 years = 4015 days = 1.11 hr for each second wasted)
Even sitting alone in a room doing nothing can be productive time. It gives you time to think about things in a way you can’t when you’re doing other activities.
There is no such thing as unusable, unproductive time unless you actively choose to make it that way.
There’s always something you can do, even if that’s reflecting on life.
What GP is saying is that your scenario is strictly worse. You meant to argue that commute time is not seen as equivalent to "lifespan time" since no one would make your trade, but in fact your trade is deleting time from a life entirely in exchange for commute whereas commute time may still otherwise be spent somewhat productively (e.g. listening to a podcast, calling your family, or whatever). The difference between someone who accepted this bargain and not, is that the person who accepted the bargain has in fact less productive time overall (because it got deleted in the bargain), so of course no one will make the trade. When I say "deleted" I'm talking about the lifespan reduction, I understand your commute becomes "instant".
edit: Nah, I was even misreading it - it's just saying exactly the same thing that objectors are saying. Commute time isn't worth nothing, and people wouldn't trade it 1:1 for time not spent commuting.
I'd comment by saying that human life math, and X% disability math etc., though absurd, has to be done when you have a say over how lives and health will be spent. There will always be death panels.
The best you can do is make your calculations explicit, so other people can express how they object to them and how they would do them differently. But there's no inherent or consistent value to a life or a handicap. Your dying would hurt you more than it would hurt me, and would affect your children in ways in which I wouldn't be affected; and losing the use of one's legs would be worse for an NBA player than for a computer programmer. Losing an index finger could be nearly like death to a guitarist.
Shit, I would. Trading a year of the lowest quality life I'll ever experience, for a couple of years worth of saved time at a higher quality of life, PLUS the rest of my time can be better quality, too, since I'd be better rested? Sign me up.
> I think not. Spending an hour waiting for, or on, a train is not the equivalent of losing an hour of your life.
I think that this is closer than you think and would be debatable. If somebody told me that a 1.5-2 hour commute five days a week would go to zero for the rest of my life, and I just had to give up the last year of it, I'd probably take that deal.
edit: I have to agree that the commute is negative time for me, then. It's worse than nothing. Yes, I've gotten a lot of reading done, but people get a lot of reading done in prison, too.
But you wouldn't be giving up a year of your life with your child. You'd get the same amount of time to spend with them - it would just be more-front-loaded.
This is a great story and it illustrates why I hate the "Trolley Problem" so much. In the real world there is always ambiguity both about the state of the world and the effect of one's actions on the state of the world.
Trying to reason about these things and apply metrics and criteria to the cost of various decisions simply cannot be made in an objective way. The subjective nature of it will outweigh any of that by a factor big enough to make all of those numbers useless.
The most common mistake is treating moral decision theory as a framework for making moral decisions, rather than what it actually is, which is a framework to try to explain moral decisions. In my mind this is the failure of rationalism as a project, in that it assumes that we know more about the world than is possible and a first-principles approach to the world is viable, and attempts to treat these frameworks as ways to guide a decision when the world is messy and does not possess an axiomatic framework even approximately. It's a descriptive / prescriptive divide, a map vs. territory divide, and it is unresolvable.
We make laws and pursue justice to try to get the extreme edges of the social contract under control, but we know that these things are always going to rely on human judgement.
Excellent trolley problem. Living with your decisions is unavoidable. I would be able to live with "I stopped a train because I wasn't sure if someone was about to be maimed."
> But since the "it's ok" position was harder to communicate, I think part of what happened was what we could call "expressivity bias", where an idea that was harder to express lost out, combined with it just being very hard to integrate diverse viewpoints under time pressure in a stressful situation.
This sounds like something a philosopher would say. I'm not a psychiatrist but individual and group psychology is more complicated. Most people tend to assume somebody else will take initiative. Also, you can show incontrovertible evidence to a crowd and they may still express apathy (e.g. the Kitty Genovese case).
> Most people tend to assume somebody else will take initiative. Also, you can show incontrovertible evidence to a crowd and they may still express apathy (e.g. the Kitty Genovese case).
For the curious, I recommend reading some recent reexaminations of the Kitty Genovese case. Most people's perception of it is as a demonstration of the bystander effect, but if you look more closely, that is not a very honest take of what happened. That angle however is something that shocks people and grabs attention, which makes it a very successful meme.
Rutget Bregman writes about this in his book Humankind, as well as other similar cases. Some snippets of it can be found online.
I don't think that this is a hard problem, if you look at it the right way: after the fact. It is the end state that is important.
Here are the end states:
Choice 1: You pulled the alarm when you believed there was reasonable cause. An ordinary person would believe the same.
------
Outcome 1: It was only a bag, so you look foolish and maybe have people annoyed that you stopped the train for nothing.
Outcome 2: The person could have damaged their arm seriously. You did a good deed and look good and maybe people are annoyed that you stopped the train.
Choice 2: You do not pull the alarm when you believed there was reasonable cause, as above.
----
Outcome 3: Only a bag so the person is annoyed that you did not and maybe other people.
Outcome 4: The person suffered severe injury. People are annoyed that you did not stop the train. You are clearly morally wrong and have lost part of your humanity.
Given that you reasonably thought the person's hand was in the door, the only ethical and moral action is to stop the train. It makes no difference what actually happened after the fact.
While the other possible outcomes may be unpleasant, the one where you stood by and let a person be hurt is the worst.
This is not an idle speculation. Once I was driving on an inter-state and a person who was seemingly drunk came zooming up, weaving in and out of traffic crossing lane lines. I could have called, highway patrol and did not.
My exit came up and after I took it I heard on the news that there had been an accident on the highway, in my direction a few minutes later. I don't know it was the drunk, but it could have been. No math, no analysis will save you from losing some of your humanity when you could have saved someone from harm and choose not to do so.
[formatted for better clarity. Hopefully]
Here I think you are trying to make this from a hard problem into an easy one, but a lot more hinges on your use of "reasonable" and "ordinary".
That then becomes the crux of this; someone who hears "hey, hold the door" and pulls the brake is entirely different world from this scenario, even if they held the belief that someone was in danger.
In the real world moral decisions are incredibly complex because they always have an element of uncertainty about the state of the world and the effect your actions will have on it.
So in the end, judging real life moral decisions always comes down to applying some standard of reasonableness; that's the hard problem and it is irreducible.
Problem is, we live in a society filled with sociopaths who do not care about injuring someone else to save 1-2 hours of their own time. Look at all the comments on this site defending Huang's affairs. Seems our society has lost any sense of morality.
I think this is not as interesting a philosophical problem as he is making it out to be. If you are standing next to the brake and someone says 'hit the brake his arm is stuck', then you hit the brake. You don't work out some complicated moral calculus.
> I think this is not as interesting a philosophical problem as he is making it out to be. If you are standing next to the brake and someone says 'hit the brake his arm is stuck', then you hit the brake.
If it's expected/uninteresting for someone to hit the brake when someone shouts at them to do so, then is it not interesting to consider why the opposite happened in this case? ("I pointed to the person next to it and shouted [...] They didn't, so I ran down the train")
> You don't work out some complicated moral calculus.
Which I think is think is one of the author's points, that "morality problems often suggest a frame of mind where we have time to fully reason out the likely effects of each of our options" and that doesn't match reality.
I don't think the author was describing it as an interesting philosophical problem; rather as the difference between the real world and philosophy class. Indeed the real-world problem is not one of moral calculus. It's reacting in a decisive & timely manner with partial information. Most of us are not well-prepared for emergencies, and philosophy class is kinda laughable as an idea of adequate preparation.
What if it results in another high speed train rear-ending yours, and dozens of deaths?
The reason you think about things ahead of time is so you don't make an instinctual or socially-pressured decision in the moment which may be bad.
In this specific case, it's whether you even have a brake available for passengers to use, or you decide that having a brake there would usually make the situation worse. Passengers are ignorant of most of the consequences, they can only react to what they see in front of them. The reason you put a brake there is because the driver would likely brake if they could see what was going on, but they can't have and process all the information about everything in the train at all times.
i.e. "If you are standing next to the brake and someone says 'hit the brake his arm is stuck'," and there's a brake for you to use, somebody has already made the decision that your ignorant decision will likely be a good one.
Anyway, somebody's arm getting torn off, a scene spread over a quarter-mile of train tunnel is certainly going to delay trains a lot more than a quick stop, and leave a bunch of people in the train traumatized, with large dry-cleaning bills. There's a way to make this into an interesting problem by changing a bunch of stuff, but it actually isn't. Not pulling the brake isn't going to avoid anything if somebody's arm is stuck in the door. The reason they put the brake there is because they've already calculated this.
Trains do not simply slam into the back of other trains. The recent incident in Spain occurred because the system that would have prevented that was broken, so they turned it off (as it would not have allowed the train to go while it was broken). This reminded the whole world of train operators why they Do Not Do That Ever Ever Ever.
Have you been on a train station and heard an announcement that a train is delayed due to a signaling fault? That is because they are not allowed to ignore a signal even if they know they are broken. The signal must be repaired, no exceptions. When you create an exception you get a train crash, like in Spain, so the rule is No Exceptions Ever.
Trying to do a cost/benefit analysis on a man with his arm stuck in a door of an accelerating train is a demonstration that Effective Altruism is more demonic than altruistic.
I think a lot of people have been quietly traumatized by propaganda around authority and consequences. There's an unlimited amount of this content on social media platforms.
If you can't quickly run the min/max on consequences for pulling a subway emergency brake, you probably shouldn't be allowed to ride the subway without some kind of exception on file.
I think that when people are really put in to these kinds of situations, you'll find that you don't do what you thought you would. (Assuming the situtation is actually new to you -- no fair if you already know.)
Which is really interesting!
You might do "better" than past you expected. You might do "worse". You might just do "different". But it will very rarely be what you expected you would do.
I know it surprised me when something similar happened to me. (But, alas, not a particularly good story.) And so I think that's worthy of discussion.
Yes, my experience and observations are that reactions to emergency situations are largely reflex or instinct. If you've had an opportunity to experience that particular kind of emergency before you have a better chance of keeping your wits, but in a surprising novel situation you don't really have time to think it out. You either act or you don't, and given time to examine the situation you will always wish you had done something at least a little differently.
It's important to keep this in mind when watching a video after an event. Chances are that as a viewer you know SOMETHING is going to happen before you even start the video, which puts you well ahead of the participants already, and of course by the time you've seen the slow-mo you can plot a better course of action than was taken at the time.
> If you can't quickly run the min/max on consequences for pulling a subway emergency brake, you probably shouldn't be allowed to ride the subway without some kind of exception on file.
> If you can't quickly run the min/max on consequences for pulling a subway emergency brake, you probably shouldn't be allowed to ride the subway without some kind of exception on file.
I am not convinced by these kinds of calculations, and I think they generally do rationalists & their like a big disfavor.
A year of someone's life is approximately 9,000 hours.
If you spend an hour commuting to work and back for 30+ years, that's approximately the same amount of time.
But would anyone accept a magic genie offer where they agree to die a year earlier, in exchange for a magic ring that lets them teleport to work and back, skipping the commute? (Hand-waving away the various abuses of that sort of power; the genie is crafty, and will absolutely not allow you to metagame this sort of thing.)
I think not. Spending an hour waiting for, or on, a train is not the equivalent of losing an hour of your life.
If my commute was in a car then yes.
EDIT: even if you added the stipulation that I must continue to live where I currently do, and can't take advantage of this teleportation to live in a more-favourable area that is further from my work.
> the genie is crafty, and will absolutely not allow you to metagame this sort of thing
Besides that, this happened 11 years ago. Probably everyone involved has collectively wasted that 4.3 years just by not optimizing their time to within 1 second per day (11 years = 4015 days = 1.11 hr for each second wasted)
The original setup was spending time in line vs having usable, productive time.
You changed it to spending time in line vs the time just vanishes.
There is no such thing as unusable, unproductive time unless you actively choose to make it that way.
There’s always something you can do, even if that’s reflecting on life.
I'd comment by saying that human life math, and X% disability math etc., though absurd, has to be done when you have a say over how lives and health will be spent. There will always be death panels.
The best you can do is make your calculations explicit, so other people can express how they object to them and how they would do them differently. But there's no inherent or consistent value to a life or a handicap. Your dying would hurt you more than it would hurt me, and would affect your children in ways in which I wouldn't be affected; and losing the use of one's legs would be worse for an NBA player than for a computer programmer. Losing an index finger could be nearly like death to a guitarist.
I spent two years driving from Maryland to Quantico every day 3-4 hours round trip because I was assigned to DC but my duty station was in Quantico.
Those two years probably took a year off my life
I think that this is closer than you think and would be debatable. If somebody told me that a 1.5-2 hour commute five days a week would go to zero for the rest of my life, and I just had to give up the last year of it, I'd probably take that deal.
edit: I have to agree that the commute is negative time for me, then. It's worse than nothing. Yes, I've gotten a lot of reading done, but people get a lot of reading done in prison, too.
Would I give up a year of my life with my child, just so I didn't have to go into the office? Absolutely not!
Trying to reason about these things and apply metrics and criteria to the cost of various decisions simply cannot be made in an objective way. The subjective nature of it will outweigh any of that by a factor big enough to make all of those numbers useless.
The most common mistake is treating moral decision theory as a framework for making moral decisions, rather than what it actually is, which is a framework to try to explain moral decisions. In my mind this is the failure of rationalism as a project, in that it assumes that we know more about the world than is possible and a first-principles approach to the world is viable, and attempts to treat these frameworks as ways to guide a decision when the world is messy and does not possess an axiomatic framework even approximately. It's a descriptive / prescriptive divide, a map vs. territory divide, and it is unresolvable.
We make laws and pursue justice to try to get the extreme edges of the social contract under control, but we know that these things are always going to rely on human judgement.
This sounds like something a philosopher would say. I'm not a psychiatrist but individual and group psychology is more complicated. Most people tend to assume somebody else will take initiative. Also, you can show incontrovertible evidence to a crowd and they may still express apathy (e.g. the Kitty Genovese case).
For the curious, I recommend reading some recent reexaminations of the Kitty Genovese case. Most people's perception of it is as a demonstration of the bystander effect, but if you look more closely, that is not a very honest take of what happened. That angle however is something that shocks people and grabs attention, which makes it a very successful meme.
Rutget Bregman writes about this in his book Humankind, as well as other similar cases. Some snippets of it can be found online.
Here are the end states:
Choice 1: You pulled the alarm when you believed there was reasonable cause. An ordinary person would believe the same. ------
Outcome 1: It was only a bag, so you look foolish and maybe have people annoyed that you stopped the train for nothing.
Outcome 2: The person could have damaged their arm seriously. You did a good deed and look good and maybe people are annoyed that you stopped the train.
Choice 2: You do not pull the alarm when you believed there was reasonable cause, as above. ----
Outcome 3: Only a bag so the person is annoyed that you did not and maybe other people.
Outcome 4: The person suffered severe injury. People are annoyed that you did not stop the train. You are clearly morally wrong and have lost part of your humanity.
Given that you reasonably thought the person's hand was in the door, the only ethical and moral action is to stop the train. It makes no difference what actually happened after the fact.
While the other possible outcomes may be unpleasant, the one where you stood by and let a person be hurt is the worst.
This is not an idle speculation. Once I was driving on an inter-state and a person who was seemingly drunk came zooming up, weaving in and out of traffic crossing lane lines. I could have called, highway patrol and did not.
My exit came up and after I took it I heard on the news that there had been an accident on the highway, in my direction a few minutes later. I don't know it was the drunk, but it could have been. No math, no analysis will save you from losing some of your humanity when you could have saved someone from harm and choose not to do so. [formatted for better clarity. Hopefully]
That then becomes the crux of this; someone who hears "hey, hold the door" and pulls the brake is entirely different world from this scenario, even if they held the belief that someone was in danger.
In the real world moral decisions are incredibly complex because they always have an element of uncertainty about the state of the world and the effect your actions will have on it.
So in the end, judging real life moral decisions always comes down to applying some standard of reasonableness; that's the hard problem and it is irreducible.
I agree though, we live in a society filled with people who wish for a second Holocaust so that gasoline will be cheaper.
If it's expected/uninteresting for someone to hit the brake when someone shouts at them to do so, then is it not interesting to consider why the opposite happened in this case? ("I pointed to the person next to it and shouted [...] They didn't, so I ran down the train")
> You don't work out some complicated moral calculus.
Which I think is think is one of the author's points, that "morality problems often suggest a frame of mind where we have time to fully reason out the likely effects of each of our options" and that doesn't match reality.
The reason you think about things ahead of time is so you don't make an instinctual or socially-pressured decision in the moment which may be bad.
In this specific case, it's whether you even have a brake available for passengers to use, or you decide that having a brake there would usually make the situation worse. Passengers are ignorant of most of the consequences, they can only react to what they see in front of them. The reason you put a brake there is because the driver would likely brake if they could see what was going on, but they can't have and process all the information about everything in the train at all times.
i.e. "If you are standing next to the brake and someone says 'hit the brake his arm is stuck'," and there's a brake for you to use, somebody has already made the decision that your ignorant decision will likely be a good one.
Anyway, somebody's arm getting torn off, a scene spread over a quarter-mile of train tunnel is certainly going to delay trains a lot more than a quick stop, and leave a bunch of people in the train traumatized, with large dry-cleaning bills. There's a way to make this into an interesting problem by changing a bunch of stuff, but it actually isn't. Not pulling the brake isn't going to avoid anything if somebody's arm is stuck in the door. The reason they put the brake there is because they've already calculated this.
Have you been on a train station and heard an announcement that a train is delayed due to a signaling fault? That is because they are not allowed to ignore a signal even if they know they are broken. The signal must be repaired, no exceptions. When you create an exception you get a train crash, like in Spain, so the rule is No Exceptions Ever.
I think a lot of people have been quietly traumatized by propaganda around authority and consequences. There's an unlimited amount of this content on social media platforms.
If you can't quickly run the min/max on consequences for pulling a subway emergency brake, you probably shouldn't be allowed to ride the subway without some kind of exception on file.
Which is really interesting!
You might do "better" than past you expected. You might do "worse". You might just do "different". But it will very rarely be what you expected you would do.
I know it surprised me when something similar happened to me. (But, alas, not a particularly good story.) And so I think that's worthy of discussion.
It's important to keep this in mind when watching a video after an event. Chances are that as a viewer you know SOMETHING is going to happen before you even start the video, which puts you well ahead of the participants already, and of course by the time you've seen the slow-mo you can plot a better course of action than was taken at the time.
I can't tell whether or not this is satire
> If you can't quickly run the min/max on consequences for pulling a subway emergency brake, you probably shouldn't be allowed to ride the subway without some kind of exception on file.
Welcome to Buridan's Ass: https://en.wikipedia.org/wiki/Buridan%27s_ass