In the coinflip example @Lenoxus posted the blue pressers (and half the red pressers) might die if the number of blue pressers is less than a few 10s of thousands.
In the original the blue pressers will definitely die if the number of blue pressers is less than 4 billion (and the red pressers all survive whatever).
So the blue press is around 5 orders of magnitude less risky in your variant. That’s not a nudge that’s a rocket ship to the other side of the solar system.
If anything it highlights how incredibly risky pressing blue is in the original dilemma. It’s not like the blue press has no risk in your example it’s just the kind of tiny risk we are used to accepting in life. Something like “there is a tiny fraction of percent chance of my dying on any given day, if I decide to go skiing that chance doubles”. The risk of a blue press in the original is like 10000x more risky than that! How many sign ups would you get if you run an extreme ski trip where the chances of dying are 10000 times greater than a regular ski trip? Not many I think, regardless of how great the apres ski is
The reason I’m so pro-blue is exactly because I expect people to wisely and sensibly avoid the death button, and instead press the life button.
Oh, you meant the “death of me, the most important person” button? Huh. Interesting. Okay. I was talking about the “needless increase in the possibility of death from a very sensible zero-death-whatsoever equilibrium” button. That button is obviously more deathy than than the blue one.
The instinct that, in my “flipped” version, you should take the antidote and congratulate yourself for your life-preserving rationality is… just nonsense. Did you assume that getting an injection would be 100% painless? Did you imagine that the antidote probably protects you from additional poisons after the experiment is over, so it’s a fun bonus?
Let’s say we limit the scope of the experiment to one person only. Why on Earth does that person choose to take an antidote followed by a poison? Is this a normal thing a person would do? Is it more normal than taking a poison followed by an antidote? They’re both equally weird behaviors.
Now let’s add a second person, a third, etc. Where the hell is any supposed risk coming from? They’re all choosing the no-deaths-and-also-no-ouchies option. They’re all being normal. Maybe the hundredth person decides to be weird, they just like needles, but that’s no problem for anyone else.
The mindset of the red pushers has quite frankly become tedious to me … but blue pushers? What nudge, what sort and how big, would get you to switch and press red?
Two broad types that I see as intertwined: information that makes you believe a vast majority are going to choose red despite your (our) default belief; and negative incentives, like having to endure significant pain or loss of function, accompanying pressing blue. Intertwined because not only do those negative incentives impact you, you appreciate that they impact the choice others will make.
I don’t think any of us would say any cost is worth it. And I suspect that the cost that makes us as individuals begin to question at all is right at the point of catastrophic collapse to few pushing blue because we start to believe that there is that much less of a chance enough others will be willing to pay the cost.
My view is that a relatively small incentive in either direction should rationally bias us strongly to it. If I knew that everyone who presses red gets $50, I’m pressing red, and that’s not because my morals are that cheap. It’s because I find it immensely hard to imagine coordinating well over 4 billion people against it – all of them volunteering to forego the $50 because of a hope we can push the vote in the optimal way, all of them knowing that they’re making a leap of trust that all of the others forego it, etc. Also, the “nice” thing about the incentive is that, once you accept the unfortunate truth that it probably secures victory for red, it does at least incentivize each life to save itself beyond that.
It’s even harder to stay strong if the cost isn’t in opportunity, but something visceral, like a finger or even a tooth. Of course I’d lose a finger to save just one life, that’s a no-brainer. But expecting 4 billion voluntarily lost fingers (for the good cause of saving the small number of people who mistakenly chose to lose a finger) is just not within the realm of plausible imagination.
And of course the same goes in the other direction. I already think blue is a solid winner when there’s no explicit reward for it, but even (especially) a strict “cold-hearted rationalist” should realize that 1 cent is enough of a coordination mechanism. Each of a set of 100 people who coordinate around blue and iterate a million times will each be approximately $10k richer – and more populous – than a group of 100 (at least, initially 100, oof) who coordinates around red. However, the world isn’t cold-hearted rationalists, so perhaps (if the people who currently say they’d press red are correct that polls don’t reflect a real life-and-death situation) it would take something on the order of $50 to make the right answer clear to everyone in general. (4 billion people will not cower in fear at an imaginary 4 billion people who cower in fear at the idea of 4 billion people who forego $50 solely to become potential killers.)
I’m sorry, I think I have lost the plot. Why does there need to be an imbalance in reward between red and blue? Why not… each person to select red gets $1000 no matter what. If you select blue, you and everyone else (regardless of individual choice) will get $1000 each if over half the people select blue, and nothing if less than half select blue.
It seems kind of stupid to select blue, right? Because literally everyone can go. “Oh. So if I select red, I get the money, and also everyone else can see how simple this choice is and get the money too just by selecting red? Thank you very much: red!”
It’s like “would you pick up the money… or would you leave the money, and hope that if over half the world also leaves the money, you get the money you left behind anyway?” Just cut out the absurdly complex alternative and just pick up the money, right?
I think that is exactly it. People are primed to self-identify as altruistic so long as they don’t actually have to be altruistic (that is, they are merely saying what they would do, not actually having to do the thing). Inject a little reality into things—impose a steep cost to altruism—and their calculus might well shift. Especially when, if you think about it, there really is no need for altruism: if everyone would just pick red (unless they genuinely want to die) there’d be no need for anyone to pick blue.
Anyway, my preferred framing of the original scenario would be: if you want to live, select red. If you want to die, select blue. Just know that if over half the world also vote that they want to die, you actually won’t get to die because it will overload the system and everyone will be forced to live despite their choice.
Functionally identical, but blue is not framed altruistically. It likewise does not presume that death would be considered universally undesirable. As I’ve been saying in my efforts to highlight how absurd this scenario is, some people might actually want to die and we should not discount such people from our moral calculus, but rather respect their agency and their dignity as humans and allow them the opportunity to die if they so choose without inserting our own faux-altruism into the mix.
I don’t know how to react when people say my variation on the problem was “wrong” and then provide their own variation on the problem. The point of my variation isn’t “The original problem is exactly like this”, the point is to get you to think along a certain line. I’m seriously asking: do you expect three lost fingers, or no?
A few different people I’ve talked to online say that yes, they expect three lost fingers. (Others, who press red in the original problem, get what I’m saying and say no, they expect zero lost fingers.) Their argument is “blue is still the potentially-die button, red isn’t.” It’s wildly irrational of them, in my view, but they have difficulty grasping the irrationality. They’re almost thinking of it like “Press blue and you definitely keep your fingers but have a one-third chance of dying”, whereas I (and others who think like me) realize the chance of dying is very, very close to zero because the penalty is a very compelling coordination device.
And, as a blue-presser, I just think the framing is the coordination device anyway, no prize needed. If you label the buttons KILL and DON’T KILL, pressing DON’T KILL is more correct. If you label them DIE and DON’T DIE, pressing DON’T DIE is more correct. This is not irrational and inconsistent on my part, this is fully rational because of how it take’s other people’s behavior into account. It’s the same deal with “Do you volunteer to take a poison if you know that the antidote is given if a majority takes the poison” and “Do you volunteer to take an antidote if you know that the poison is given to all 1000 people in the room if a majority take the antidote”. The correct answer in both cases is “You don’t have to do the thing, so don’t.”
Change the first one to “an unassailable majority,” and it’s right: I won’t choose blue if I think it won’t make a difference. At least, I think I would: the idea of living in a world in which the most altruistic people have all died off is pretty fucking depressing, and I’m not sure it’s a world I’m willing to inhabit..
The second one? Absolutely.
Other types: a positive incentive for pushing red, such that I think it’s going to lead to an unassailable majority for red. (“Want to live” is a specious argument, since 50%+1 blue = everyone lives, and if I want to live and also want to live in a world of altruistic people, that’s already incentive for blue).
Another system that would work:
A chance to discuss ahead of time that leads to the unassailable red majority
Failsafe measures for incompetent people – children, adults with cognitive disorders, etc. – who might be unable to make a reasonable choice.
If we can somehow guarantee that everyone will choose red (and I don’t mean “most people,” I mean literally everyone, with the only possible exception being people who meet criteria for right-to-die decisions, and even then I’d rather those decisions not muddy the waters of this one event), then I’ll be satisfied with the red choice.
Emphasis added. I think this is genuinely morally abhorrent. Akin to present bans on euthanasia or assisted suicide that center their analysis on the supposed sanctity of life while completely discounting the agency of people who want to die (whether terminally ill or not).
ETA: I originally threw abortion into the mix as well because the argument also seems to mirror the “whole life” movement (I believe that’s how they self-identify). They say they want to go beyond abortion bans to also make life more livable for children and people who have children, but funny enough they only ever seem to get as far as joining the so-called pro-life crowd in pushing abortion bans. Whether the desire to look beyond the instant issue and take care of people later is genuine, the outcome only ever seems to be that people who are already marginalized become further marginalized as we once more discount them from our moral calculus. Anyway, since it needed a little more explaining, I’ve pulled it out and put it as a standalone in the ETA.
I’m specifically not discussing the morality of assisted suicide here. I’m saying that if assisted suicide is appropriate under certain circumstances, there are already plenty of ways to enact it; using the buttons to enact it muddies the question (who is in a mental state to make the choice? Is a seven-year-old? Is someone suffering from a depressive phase of bipolar disorder? Is a fifteen-year-old suffering from his first heartbreak? How are we using the buttons to make these nice moral distinctions?)
Unless you think that anyone at any time, whether suffering from depression or being seven years old or being a heartbroken teenager, should have full and impulsive access to the means of suicide at the literal press of a button, I think it’s morally abhorrent to use the buttons in this way.
Yes! Because that’s how human brains work. Everyone does think they are the most important person. That’s not to say they are self serving automatons they can be selfless and altruistic but but fundamentally we think we are the most important person.
That’s why a small nudge is not going to change the odds very much, we aren’t asking people to give up some of their prize money in a Game Theory thought experiment, we are asking them risk a significant chance of instant death. A few bucks or a different wording is not going to make that any more attractive.
I don’t this holds in reverse though. No amount of nudging will convince someone to risk death, I don’t think it would take much to nudge someone into not risking death. IMO most blue pushers would seize on the fifty bucks as a reason not to risk their life, I mean if someone is paying you fifty bucks to press red there there must be a reason to right? .
Well that requirement is the bit that I find interesting.
Excuse if the following meanders …
I thought of this thread while reading the “stocking up” thread where hoarding was under discussion. As long as most believe/trust that most others aren’t going to hoard, then few will hoard, shortages don’t occur, everyone has enough. But once we individually perceive that others are buying up months worth of toilet paper out of fear there won’t be enough to go around we en masse do the same, and some of us are literally SOL. We don’t have to know that an unassailable majority of others are buying up all they can; we just have to see it … enough. Then we want to buy as much TP as we can before the hoarders do. It just takes enough that we don’t trust enough will continue to rationally act in manner that best serves the whole. Then it is a tipping point and fights break out over the last roll on the shelf!
My WAG is that there is a similar tipping point for most of us blue button pushers. I doubt it is just a dollar. But I also doubt it is huge.
It is exactly like hoarding, yes! That’s a perfect parallel. Red-pressers are, in a sense, contradictory if they don’t engage in hoarding everything they can right now (or in bank runs if those still existed). They intuit “Why, absent a crisis like covid, would half of my town suddenly hoard toilet paper?” and yet don’t apply that here. They think a vast majority of humanity are one “Boo!” away from a dangerous stampede.
No amount of nudging will convince someone to risk death
What if you can only press red after first pulling out one of your own teeth with provided pliers? We first filter the set of button-pressers to able-bodied adults, etc. Do you still do it? Do you not see the absurdity of doing it?
It isn’t, and maybe the Kantian imperative is helpful here.
I’m definitely not a professional philosopher, and my understanding of the Kantian Imperative comes from a book I read when I was like 17. But my understanding is that it roughly says, “It’s okay to do something if theoretically everyone could do that thing and have similar results.” So, it’s okay for you to take a walk after dinner, because everyone could take a walk after dinner. But it’s no okay for you to shoplift, because if everyone shoplifted, the stores would shut down, and then you couldn’t shoplift.
Apply this to red-button. The argument goes that it’s okay for you to push the red button, because everyone could push the red button. (It’s also okay, for similar reasons, to push the blue button, so Kant’s pretty useless here). But apply it to hoarding: can everyone hoard? Almost by definition, no. So it’s not okay, and it’s different from red-buttoning.
Though that’s not a nudge! That’s a major change to the setup of the thought experiment.
Yeah, you can add a huge negative consequence to pressing red or a huge positive consequence to pressing blue, and it might change the odds. But it needs to be a big enough to convince your average person to risk a significant chance of death. A little nudge isn’t going to do it.
Again what’s absurd about it? It’s just accepting a fundamental thing about humans that they don’t like to die. It will take a lot to convince your average person to risk death. It’s not saying humans are some kind of selfish automatons that will never be altruistic, just that they don’t like to die
I don’t really think these kind of prisoners dilemma thought experiments are that applicable to real life. But if they are that’s hardly an argument against pressing red. Have you seen a supermarket before a big storm? Everyone absolutely “presses red”.
They absolutely still exist and everyone still absolutely presses red..
Not a philosopher either but personally if that is an accurate summary of the Kantian imperative then I’ll stick to a more utilitarian perspective.
I am in a crowded room with one exit. Is it okay for me to get up and leave at this precise moment? Well if everyone did the same, got up to leave at that moment, it would be disaster! Since everyone can’t do it then I should stay seated?
I’m looking for best possible outcomes (for both myself and the group overall) given an internalized understanding of reality and the probable behaviors of others. If my understanding of that reality changes, my choice changes as well.
Not necessarily! Say there’s an old dude with a walker who’s trying to make it through the door: of course it’s not okay for you to push past him. But it’s okay for you to get up when there’s space, to go to the back of any line that exists of people trying to leave, and to leave when it’s your turn. Which is probably what you’d do anyway, and that’s a principle that everyone in the room could follow.
