This is the focus of the Anscombe-Foot discussion. Ethics is not algorithmic.
Indeed, very little is algorithmic, including maths and logic.
So some—many—when confronted with the “trolley” problem presume that there is a deducible correct answer and argue accordingly.
Others—the better thinkers, I will hold—will see the Tram problem; that there is no optimal solution, yet nevertheless we must act.
The red voters see the trolley problem. The blue voters might be seeing the tram problem.
Here’s another way to think about it. Suppose we really were put in the position described, perhaps by a victorious Kim Jong Un or some such.
The simplest way to give him the finger would be to refuse participation. But if that is not an option, vote blue so as to maintain the status quo and so undermine the very problem as offered.
Voting blue as recognising the framing and rejecting the tyranny.
@Moliere ?
to the whole thing?
Damnit, I’ve spent too long on this silly thread.
Me too.
But, to be fair, I’ve been having a hard time saying something at all, so it brought me out even if just to complain and gripe 
That’s the way I’m inclined just because it seems like the only way to answer is to bring your presuppositions to it.
What you’ve said of bluepushers fits my ethical intuitions. It’s not necessarily a reasonable proposition, but it is at the same time quite reasonable, given what we know about human beings.
I’m inclined to throw the finger mostly because it looks like a party question which is fun, but then we can spend too much time on the thread (as I have too)
That’s not my argument.
Your argument is that you should vote blue, but if you have reason to believe most will vote red, you don’t have to tell anyone how you’d vote because it will acknowledge how you vote is contingent upon this facts.
If most vote blue, my argument is that how you vote is morally and factually irrelevant because the conclusion is determined and your vote has no effect.
Glad to hear it.
What you suggest is not my argument, either.
A slide from"my vote probably won’t determine the outcome" to “my vote therefore does not matter”. It’s not just invalid, it’s downright stupid. Pretty sad to see you offer it.
It’s nearly a week since I said the same thing:
Shortly after that, I offered the one (red) button version of the game, which I think is better: here the question is whether you can resist the urge the push the red button. (You can do it the other way too, with just the blue button.)
I compared resisting the urge to push the button to the marshmallow test, and that’s actually kind of interesting, because the marshmallow test probably (this kind of stuff is never cut and dried) turns out not to be tracking what it thought it was, but something more like the stability and trustworthiness of the subject’s world. Have adults reliably provided for you? The nice lady promised to give you another marshmallow—have adults generally kept their promises to you? And so on.
So the marshmallow test probably wasn’t really about self-control but about trust, so maybe my comparison still holds. Or, rather, the kind of self-control that makes so many things possible is itself only possible given the relative safety and stability of the social order we have fashioned, or only to the extent that we have made the world more trustworthy.
My vote doesn’t matter if 51% have voted blue. That’s not stupid. It’s true.
It’s the opposite of having 51% voted red. Your vote truly matters then.
And you know this, prior to voting?
For the blue voters, isn’t the moral principle advanced not “thou ought vote blue,” but instead “thou shall vote blue”?
As in, the option to vote red shouldn’t be available, or at least a serious detriment ought accompany a red vote to reduce it.
I know it based upon the best information available. That’s my point. Altering one’s vote based upon increasing overall survival is reasonable, and not being omniscient, I have to base my decisions upon my best information.
If the world were Norway, my vote wouldn’t matter. If the world were Alabama, I’m voting red.
I’ve been chatting to various AIs concerning the geography of the red/blue problem. It started by my asking for examples from India, China, or SE Asia; so far as I can see, there are none. Now that in itself is striking—that it should receive millions of views in the West and yet garner hardly any interest in the East.
Deepseek described it as WEIRD (Western, Educated, Industrialized, Rich, and Democratic).
Claude pointed out the irony that the dilemma, while supposedly about universal human nature, instead lands primarily in individualist cultures, suggesting it’s actually measuring something culturally specific.
So, a research topic for you: Are there versions in Asia, Africa, or South America? What were the results, and how and why do they differ?
Correct. Everyone votes red and everyone survives.
There are two ways to interpret the problem:
This is why I want to address the problem in a way that doesn’t depend on either of these interpretations.
All we need to ask is if the below argument is valid or not:
I say it is invalid. What do you say?
If it is invalid then the only things we need to consider are:
Therefore, I vote red. There is a rational reason to prefer voting red over blue and no ethical reason to prefer voting blue over red.
It depends on how you want to see people. If you want to see them as cynical opportunists, sure convince them that voting red makes sense.
People who vote blue are idiots.
I prefer to appeal to the good sense people in fact have (online trolling aside) that we should treat each other with respect and care.
I see no benefit in trying to argue that if we get lots of people voting red, then lots will survive, but at large number of people will die.
I think this shows how messed up we currently are. Altruism is viewed as weakness or stupidity.
In any case, this has gone on far too long for my liking.
By all means have the final word.
Every society is a rational society. But rationalities take many different forms, and no cultural system of reason can be said to be MORE rational than any other. The vote in the game does a beautiful job of revealing this multiplicity of rationalities, each faction insisting it is the ‘correct’ version. Banno’s version appeals to our better natures, you focus on personal responsibility and guilt, and @AlveK dresses up his version in the most detailed mathematics.
My version is the correct one for me, but I wouldn’t expect it to make sense from another cultural perspective.
There is something to that.
But the near universal (if not an actual universal) is to be empathetic, welcoming and treat others nicely.
It’s today’s hyper-capitalistic, super competitive dog-eat-dog world that things like these cause debate. It should not.
The protection agains
Capitalism didn’t make people competitive. It takes advantage of it. Do you think we’d be talking right now over the internet without capitalistic incentives, as if maybe through massive collaboration the internet would have sprung up, each technician n working just for the satisfaction of doing what is best for society.
This is my response to all those who mentioned me in this thread, so @Suny, @Banno and just now, @Joshs.
You’re right I made a massive mistake here. Of course it’s true that P(\text{survive}) = 1, but the absolutely probability is not the right measure, it should be the difference in the survival probability:
And that changes everything…
I’m quite busy these days, but I had a window of time where I could respond to this thread. As such, I rushed my answer too much and wound up with some fatal flaws in my mathematics.
So, I wasn’t trying to assume the Principle of Indifference. My integrals did assume the Principle of Indifference, but my actual assumption was this:
Any reasonable non-uniform distribution of likely P_r-values will not be enough to sufficiently bias them for the bluey interval’s contribution to matter…
That assumption was probably correct given my incorrect mathematics…
But after fixing my EV mistake above, I now see that the distribution of possible P_r-values is very important. The more density to the right, the better redding becomes. The more density at 0.5, the better blueing becomes. The more density to the left, assuming an okay density at 0.5, the better blueing becomes.
If it’s a multimodal distribution, high density to the left or at 0.5 might not be enough… For example, if it has a symmetric, somewhat polarized multimodal distribution, redding wins, because the hump on the right matters far more than the hump on the left…
A multimodal distribution would only arise given that we’d combine the evidence from different, competing models… which might be relevant here. Although we could just combine all the questionnaires on specifically the blue/red dilemma, see the total vote, and just model the likelihood of different P_r-values unimodally, I think that people will find this insufficient.
I introduce the idea of C_{br} and C_{rb}, which are the conversion rates of “professed bluers/redders” to “actual redders/bluers”, respectively.
Since we cannot know the latter, we cannot know what these conversion rates are, but we do know they exist… Because of course, some people judge themselves incorrectly… Especially on matters of life and death.
People could construct models taking data from comparable setups, where similar, hypothetical dilemmas became real dilemmas. I don’t think it’s a controversial claim to suggest that C_{br} > C_{rb}, because people usually underestimate their selfishness more than they understimate their altruism.
Which means, even if all our polls give us P_r, a model that takes into account some pair of conversion rates it has calculated, we’d be looking at:
But of course, there’d be a lot of debate regarding what the best model here is, and if we’d apply the Principle of Indifference to all plausible models, we could wind up with a multimodal distribution of what values of \text{model}(P_r) are likely. This complicates matters.
I want to give this topic the care that it deserves, which I don’t really have the time for nowadays… (my chance was yesterday, but I over-simplified and made some errors).
But now, I cannot say I agree with the crux of my previous posts anymore.
My position has shifted from being a redder to mostly withholding my assent. I need to figure out two things before I make up my mind.
I need to figure out the mathematics properly, as the above stemmed from a quick investigation after I realized the P(\text{survive}) mistake, so I’m not fully sure about it.
I need to see the polling numbers as they change and become statistically significant*. Preferably not just for TPF, since TPF is not even remotely a representative slice of humanity. I imagine someone will eventually set up a bigger poll somewhere on the internet, but even that could have some representativity issues since roughly 20 to 30 percent of the world does not have internet and would thus not participate…
*In the Insulated Game however, I would not have access to 2., which means once I’ve figured out 1., I could give my mathematically-reflected, but not statistically-informed view, a view that I could plausibly have/attain in the Insulated Game.
But ending with the above would be a cop-out… Because what if I got hauled away to the red/blue game right now? I would not be able to work out the mathematics in time (unless they’d give me unlimited time in there).
To withhold one’s assent is merely a privilege of circumstance… So, if I was forced to make a decision right now, what would I do? Well, given that my current tentative mathematical understanding is such that even in a case where our estimate for what P_r is a multimodal distribution concentrated on the left, even just a small hump to the right could be enough to trigger a very aggressive term in the EV formulae, thus making redding best.
I don’t think the density on 0.5 is great enough, and/or the right-side density low enough, that blueing has the highest EV. I also have a very strong survival instinct and you can call me selfish, and maybe I am, but given my current partial and tentative understanding, I believe the EV of redding is still best.
But now it’s actually a matter of how the data updates, for me. Earlier, I thought no plausible data set could possibly make blueing the highest EV choice, but that was based on my mathematical errors. Now, I’m genuinely unsure. Thank you @Suny for pointing out the errors! I hope I’ll have time to return to this thread so I can set up the proper EV formulae, maybe set up a Python script to figure out some stuff, and then also look at what the statistics say at that point. Maybe make a good model for what we can expect P_r to be.
Yeah. It’s perfectly conceivable and in fact that’s what happened in this world.
The internet was not developed by private business, it was built for military purposes, then got privatized.
You only hear the success story after it went private.
Evil is what you define it as , if you were bought up being told you are your main priority saving yourself wouldn’t be considered evil in your eyes . Another argument is that , the person simply assumes everyone will opt for the red button as it is the safest option but ofcourse children and the mentally imprared and the dumb are expection to this but as everyone will die at some point either way this event make sure that only the best of the human species moves on from this event as the event on a whole is unavoidable.