To blue or not to blue?

Ethics
122

It’s the same as in the OP. If you push the blue button and most don’t then you die. I’m simply showing that there are two ways to interpret the problem:

  1. Pushing the red button means “save me” and pushing the blue button means “save me only if most push the blue button”
  2. Pushing the red button means “only kill those who push the blue button” and pushing the blue button means “don’t kill anyone”

Those who say that it’s unethical to push the red button interpret the problem as (2), but that interpretation is not necessarily the correct one. The OP only says that if most push the blue button then everyone survives and that if most push the red button then only reds survive, but these material implications do not entail (2) given that they are also true if (1) is true.

123

Cool. And my point isn’t really about being judged, its about feeling like you can judge others based on how they chose buttons.

My quarrel with situation ethics in general is that you have to import too many facts to really infer much about the person who pulls the level or pushes the button.

And further, two of these facts that are absent are the key ingredients of morality - intent, and willing consent of the agent - without these, there is no moral act at all.

So you can easily, quickly diffuse these hypos by admitting at the start that the situation is not even a moral question. Blue or red never imply intent or or consent - so both button choices can always be morally justified depending on how you import those facts.

Before we can ask a moral question, it matters whether I am being forced into this preposterous situation or not. It matters whether the situation is preposterous or not. It matters what I know and what I intend to be the outcome of my action. Without inserting facts about these into the situation, zero morality can be inferred.

Like here:

That’s a logic game. The only thing that you need to assume is that people don’t want to die by gas chamber. You don’t even have to assume murder is immoral to give an answer. If you can reason, everyone pushes the red button. I thought the OP prompt was seeing who could reason their way through to that logical conclusion.

Situation ethics can be a way to raise the topic of morality, but morality remains another topic. And Sit ethics is, in my view, a terrible way to determine what is the moral choice and how to make a moral choice (how to live morally). The only plain discussion of morality here is, what’s up with the guy who built a world where everyone can be murdered by pressing a button selecting on probability? Was he moral to even pose this question to anyone, let alone attempt to enlist moral agents in whatever he was planning to do?

So many around here jump to “red button pushers are murderers” and “blue button pushers are good people” - both are false, on the facts given, and neither conclusion can be inferred simply from the nature of what the red or blue buttons do.

I think we all have to say that the guy who built the less than 50% blue button pushing killing device is solely responsible for who dies. And further, it would be immoral to hold anyone else responsible, because they were forced to push a button with death looming everywhere.

But so what - we still don’t know what morality is. Why is it immoral for the guy to build a little experiment? We all die anyway, why not, in the Blood Meridian type of spirit, up the stakes for merely drawing your next breath. We should have to vote every morning before we wake up, to make life more interesting, or you are immoral for being boring.

I know, I’m just not playing now, but like I said, I must be a moron, because when I first answered, I thought we were playing dice with what makes sense given the psychology and smarts of other button pushers - not what makes sense given the moral values of other people. I played assuming most people all want to live and don’t want anyone to be murdered by anyone. You smarter people were all seeing if I would implicate myself in a mass murder.

124

There are two related questions:

  1. Which is the most rational choice?
  2. Which is the most ethical choice?

Given that it’s a non-zero sum game and that everyone can guarantee their own survival by pushing the red button, from a game-theoretic standpoint the most rational choice is to push the red button, and only the irrational would push the blue button.

However, it might be that pushing the red button is the least ethical choice, e.g. if pushing the red button means “kill everyone who pushes the blue button” and some kind of deontology is correct. But as the OP doesn’t specify this, no answer can be given.

125

If I help Bob with the project, I might save him from getting fired, but I also might get fired myself if things don’t go well, but if I keep to myself, I won’t get fired.

That seems to ask the same question without interjecting any moral element into it because Bob getting fired isn’t like getting killed.

Assuming your goal is not getting fired, you don’t help Bob. If your goal is being a friend to Bob, then help him. If you have other obligations (like to your wife and kids), then you have to consider those when you decide to take risks with your job.

And now I apply it to the OP, which is the same. I don’t jeopardize my life as if it’s just my own. That, in itself, is selfish. There is a dehumanized element to this commentary that fails to appreciate that life itself is sacred and the one who possesses it does not have special status to jeopardize it.

If I have your gold and I am faced with securing it or gambling it and doubling it, what is the right thing to do? And by “your” in this example references something beyond me. There is a stewardship here that I don’t think is being appreciated when it comes to life as opposed to an ownership.

126

It’s not the same, because Bob not being fired depends entirely on your choice. He does not have the power to save himself, whereas in the OP everyone has the power to save themselves. That’s an essential difference.

A better example is: John, Jim, and Jane can all either do their own jobs — that they can complete without help — or help the others. If John and Jim choose to do their own jobs and Jane chooses to help them then Jane is fired because her job isn’t completed. She only has herself to blame. Her “charity” is unnecessary and foolish.

Choosing to help others is the very thing that creates the need for others to help you. That’s the true (unethical) selfishness. If everybody just does their own damn job then nobody needs to risk their career and if everybody just pushes the damn red button then nobody needs to risk their life.

127

Yes, and because the OP doesn’t specify this, the only way to make a choice is to make a logical determination.

It’s not that the OP presents a moral question or issue that can’t be resolved, it’s that it doesn’t present a moral question at all.

There is no “most ethical choice” between the two buttons. Ethics can be shoe-horned into the analysis by choosing not to play the sick game-designer’s game at all, or by judging the game-designer as the one with all moral responsibility. Ethics remains outside the four-corners of the scenario, and ethical questions don’t arise inside the game.

128

I would just add a wrinkle here. Pressing the red button in the OP’s scenario marginally decreases the chances of saving blue button pushers, whereas pushing the blue button marginally increases the chances of saving fellow blue button pushers. How many of the button pushers recognize this is an open question, but perhaps it is this marginal power to influence the outcome for others which is the basis for the moralizing going on in the comments.

129

This little logical puzzles are for Philosophy what experimental psychology is for the human intelligence deployed by great literature. No risk this will arm the devastation of the world by money moguls…

130

I’ll share some thoughts on the problem. I, for one, like the problem as it pertains to how framing can influence how we view presumably equivalent situations.

I am not sure I find the highly contextualized scenarios convincing. There is the blender thing to favor red but we can find one that favors blue. One goes like this:

“It’s an election between the fascist party and the B party. If the fascists win, they’ll kill the non-fascists and if the B party wins, nobody gets killed. Non-fascists are anyone who voted for the B party. Who should you vote for?”

And here, red would be voting for the fascists and “just stay out of the blender” becomes “just vote for the fascists”. I don’t really want to discuss if this particular version represents the problem “correctly”. That’s not the point; my point is that in both cases, the scenarios add non-trivial context to push you towards a certain outcome.

I’ll say that for people who boast about red being the obvious rational choice, they do very little analysis. Let’s try to do just that.

I’ll map it like a prisoner’s dilemma. You don’t know what the others will decide but you can’t influence that so you can already lay out three possibilities and the outcomes:

  1. Blue will lose even if you vote blue
  2. Blue wins iff you vote blue
  3. Blue wins regardless of your vote

Let’s say there are 2X+1 participants (an odd number to simplify things). We get this table:

Less than X voted blue exactly X voted blue More than X voted blue
Blue -1 + X 0
Red +1 - X 0
Explanations

One good thing to note is that your choice is causally independent from the cases outlined.

If not enough people voted blue without you, it’s obvious that if you vote blue, you are not saving anyone and in fact only dooming yourself compared to if you had voted red. If it is already the case that more people voted blue then your vote doesn’t matter.

If you are the deciding vote, then the first thing to realize is that you survive whatever you vote for. The only difference is that voting blue saves the other participants too.

It’s important to realize we are not saying anything about you causing the death of the others or being responsible for it, it’s just the difference in what happens in each scenario.

This is already taking some assumptions. For example, I am assuming that you value your life the same way you value the life of any individual. But to amend this, you would just add a coefficient c in the first column (Everyone else counts for 1 and you count for c, presumably c > 1).

The first thing to realize is that unlike some problems where it can be said that the rational solution is obvious, here no option dominates the other. This means that it is not true that whatever case you are in, one solution is always better than the other. So, what’s left to do is to assign probability to the cases and estimate the expected value of each choice.

These probabilities depend on how you believe people will behave. One simple but, in my opinion, flawed way to do things, is to take Michael’s way of doing:

I don’t think it’s true that one ought to consider it a coin toss but we can see what happens if we consider it a coin toss.

So, assuming everyone has 50% chance of voting blue, what is the probability of less than X participants voting blue? Well the simplest is to first compute the probability that we get exactly X participants voting blue, it is:

P_m = \frac{\binom{2X}{X}}{2^{2X}}
Explanations

Basically, we have 2X coin tosses. The total number of outcomes is 2^{2X}. Each coin can either give blue or red so 2 possibilities and you have 2X coins.

Now we want the outcomes with exactly X blue. This means we must choose X positions out of the 2X tosses to be heads (the remaining X will automatically be tails). The number of ways to do this is given by the binomial coefficient: \binom{2X}{X}

The interesting thing about this formula is that it decreases when X increases. So, we have that and the probability of having less than X blue votes is the same as the probability of having more than X blue votes so it’s:

P_d = \frac{1 - P_m}{2}

So, the expected values of voting blue and red (assuming others vote randomly) are:

E_b = -P_d + X P_m = \frac{P_m - 1 + 2XP_m}{2} = \frac{(2X+1)P_m - 1}{2}\\ E_r = P_d - X P_m = \frac{1 - P_m - 2XP_m}{2} = \frac{1 - (2X+1)P_m}{2}

Great, what do we do now? We try to find for which values of X is E_b greater than E_r. First thing we can see is that if (2X+1)P_m is greater than 1 then E_b will be positive and E_r negative so E_b will be greater than E_r. And I won’t go into details, but it is always true that (2X+1)P_m \geq 1.

We can check for some values.
If X = 1 (so 3 participants), P_m = 0.5 so 3 \times 0.5 = 1.5 > 1.
If X = 2 (so 5 participants), P_m = 0.375 so 5 \times 0.375 = 1.875 > 1.

So, it is always the case that the expected value of voting blue is greater than that of voting red if you assume other participants have 50% chance of picking blue. So blue is the rational choice. The reason is that there is already a non-negligible chance, your vote doesn’t matter (more blue) and when it does matter, the amount of people you could save if you are the deciding vote weights far more, despite the fact that you are probably not the deciding vote.

Now what if we don’t believe people have a 50% chance of picking blue, how does that change the analysis? The math gets a bit more complicated and I’ll perhaps touch on that in a following post but this is enough for now.

131

I don’t think you’ve done so. This is mapping it like a prisoner’s dilemma:

  1. If most push the red button then only those who pushed the red button are released from prison
  2. If most push the blue button then everyone is released from prison

Which is equivalent to this:

  1. Those who push the red button are released from prison
  2. Those who push the blue button are released from prison only if most push the blue button

The second phrasing shows that there’s no good reason for anyone to push the blue button. Everyone should push the red button, and in doing so everyone is released from prison. The first phrasing simply misleads some into taking an unnecessary risk. I’m neither obligated nor willing to risk myself to save them from themselves.

132

I thought the point of the thought experiment was whether brilliant people like you and me owed a duty to morons, too stupid to know shit from Shinola. If everyone was smart enough, of course they’d all push red, but some aren’t, so we have to think of them. And then there’s the sanctimonious people we have to save from themselves who will vastly over-estimate the percent of dumb fucks and so they’ll all vote blue thinking they’re Mother Theresas. We’ve got to save the true shit-for-brains folks and those that have decided to jump in the snake pit trying to save them.

So if you think the essential difference is that Bob is different from the typical blueguy in that Bob is truly inept but the other blueguys aren’t, then you’ve over-estimated the blueguys, which better explains why you’ve argued as you have.

133

The difference between the OP problem and @Michael 's virus example is simply that in the original example red’s survival is parasitic on most folk voting blue, while in the virus example red’s survival is guaranteed. This difference does not show up in the game-theoretical calculation because it is an ethical difference.

That’s much closer to Philippa Foot’s point in the paper in which tram problems were first mentioned. The point is that once you start to calculate, there’s a good chance you have already gone astray ethically.

Consequentialist and game-theoretical contexts abstract away from the human condition that is central to making these issues ethical problems. Ethics is about how people relate to each other, not about calculated self-interest.

If you are not in the USA, it seems you can rely on your fellows to vote blue.

Sad, but that seems to be how things are.

Note that the population here in the forum is dominated by folk from the USA, and is predominantly male. My suspicion is that males will be less likely to consider others in each scenario - that’s what the research indicates. As things stand, despite that bias, blue is slightly ahead.

So press blue. It’s the right thing to do. Solidarity, not distrust and self-preservation.

134

Also, these game-theoretical calculations assume a perfect, idealized world, and quickly break down once one introduces an imperfect, non-ideal world.

Even if game-theoretically one can justify pressing red if you somehow argue that everyone will reason like you and thus will also press red, it does not hold up once you introduce a factor where some people will press red and some people will press blue no matter what you do, or you introduce a factor where some people do not know about the vote but will be affected anyways.

135

If I know more than 50% will vote red, I should vote blue?

136

But how do you know that? That is an assumption on your part you make to justify your own actions to yourself.

137

As I have stated, given an imperfect world only pressing blue is universalizable; pressing red is only universalizable in a perfect world. As the world isn’t perfect, almost everyone pressing red cannot be justified as there is no way to guarantee everyone presses red — whereas almost everyone pressing blue can be justified because the limited number of people pressing red won’t outnumber those pressing blue if most people behave in a universalizable fashion.

138

Why does the use of calculation here only involve self-interest? Don’t we first determine how we prioritize what is valuable to us and then use calculations to help weigh options? For instance, prioritizing the needs of the group over the individual makes use of quantification to make an ethical choice. If we know nothing else about a group of people, then we may use number to help us decide. If we have a choice of helping these ten people over here vs those 10,000 people over there, and we can’t do both, our calculation determines what to do.

And the assignment of probabilities enhances the validity of the numeric calculation. If there is a reliably determinable 90% chance I will fail to help the group of 10,000 and a 90% chance of succeeding in helping the group of ten people, then I should integrate the strictly numeric information with the knowledge of probability.

139

The two phrasings are equivalent and I am doing neither or them really. I formalized simply the outcomes which are the same.

Sure. The analysis simply shows that under certain assumptions (you want to save the most amount of people, you value your life the same way you value the lives of others and you think people have 50% chance of picking blue), blue is the rational choice.

140

But isn’t it the case that your voting blue only increases the chances of saving all the people by a marginal amount, and voting red only decreases their chances by the same marginal amount, while at the same time the risk of one particular person surviving (which happens to be you) may be 50% if you vote blue but 100% if you vote red?

I sketched out a related scenario where red could be argued to be the better choice, in which the motive was not self-interest:

You are a single parent with two children. One is at home and the other is with a large group stranded in shark infested waters. You know two things. First, you have been told if enough people volunteer to come to the aid of the stranded group, it is guaranteed that this will scare away the sharks.

Second, you estimate that the odds that enough people will volunteer are no more than 50%. Therefore, your volunteering will raise the odds your child and the rest of the group will survive only by a tiny fraction, and your deciding not to volunteer will lower those odds by just as tiny a fraction. Therefore, the odds you will die trying to rescue your child are 50%. Is it ethical for you to take this chance, knowing that your other child at home will be orphaned if you die trying to save the stranded child, especially considering that your contribution will change the odds of rescue only slightly but will put you at 50% risk of death?

141

Let’s assume it is beneficial not to die. In order for red to produce the exact same benefit as blue, you need 100% of people to pick it. By contrast, blue only requires 50% of people for the same benefit. So, from a game theoretic perspective, it isn’t obvious that red is superior. It really depends on how you weigh your own survival versus others in terms of utility payouts.

Also, we could consider that surviving in a world exclusively populated by red button pushers might be inferior to glorious button martyrdom. :grin:

2 Likes