#philosophy #you lost me at the math im sorry #i want to understand #i guess i just dk where any of the numbers used are coming from. like #is the ethical median like if you asked a randomized sample of people across the world if something was moral? #bc even as a theoretical thing that would be hard to do because of cultural differences imo
The traditional formulation of ethical median maximalism concerns itself solely with the ethical average and not the source numbers from which that average should derive. One of the most compelling philosophical arguments against ethical median maximalism as traditionally formulated, in fact, is the idea that as the number of things we should or shouldn't do goes to infinity, the median value of our ethical choice might diverge.
(For those shaky on the math, that means that it's easy to decide the ethical median if there's only like three things you're considering doing. Maybe you're considering killing someone, but you've tentatively concluded that might be unethical. Then there's hugging a puppy, which you're pretty sure is okay. Then maybe you're like thinking about pirating software. As soon as you figure out whether it's okay to pirate software, you have the ethical median: two things are okay, so everything is allowed; or, two things are wrong, so nothing is okay whatsoever and you can only notionally curl up in a notional ball and modally wail. Nominally speaking, no matter how many more things you add to the list, as long as you stick to an odd number of ethical conundra, you will never risk ethical uncertainty under ethical median maximalism. However, it's possible that as you start to throw literally every possible ethical question in there, instead of a rapid convergence on a defined answer like "yeah, go for it" or "nope, we live in a fallen world or something, this stuff is bad," you will see a rapid wobbling back and forth, a "hyperexponential ethical oscillation" (which is really just technical jargon for "uh oh,") between the definite conclusion that you should and the result that you should not. This kind of ethical oscillation means that you can't turn to the already somewhat philosophically suspicious idea of mathematical limits in order to determine what the result will be when you finish throwing all possible ethical choices into the pot---it could go either way, stranding us on the tidal island of the human condition again. Ethical median maximalists generally believe that a sufficiently large sample of ethical answers will show a sufficiently firm trend towards "do" or "do not," at which point they can stop taking samples ... but of course, such ideas are not reliable.)
Why, then, do we care at all about ethical median maximalism?
Mostly, the ethical median maximalist would say, because if we do not continuously seek the ethical average, we will be forced to consider situational nuance and the possibility that doing the thing could be both good and bad, or that both doing and not doing could have merits, which is, first, functionally unsustainable, and, worse, leads to potential hyperexponential ethical oscillation itself.
What you're referring to with your thought is actually something called a randomized sampling ethical oracle, which relies on the so-called "wisdom of crowds" to resolve ethical dilemmas. As you point out, it's enormously difficult to avoid intrinsic sampling bias when you're doing this, even in the abstract thought experiment stages before you get to the point of actually trying to interview a selection of random individuals. Worse, any given sample can only solve a single ethical dilemma, as if you try to solve all ethical dilemmas simultaneously using this method you will be stuck interviewing people basically forever, after which time it will be too late to do anything ethical anyway.
As another correspondent also mentioned issues with the math, I should further clarify that "epsilon" and "delta" here are meant to represent arbitrarily small numbers.
Let me step away from the specific context of ethical median maximalism for a moment, and, for that matter, of ethics entirely, because there's a general decision-making principle here that's potentially interesting even to people who find the above discussion functionally meaningless.
Let's say you find yourself guessing about what to do.
You think ... maybe ... there's a 30% chance you should do something.
And you do it, and you turn out to be right. It was a good move.
You shouldn't immediately go back and change how you think to say "oh, I should always definitely do that kind of thing."
Like, you're baking. And you think, "I should add more salt? Maybe?"
And you go "oh what the heck," and do it.
And it comes out well. That's confirmation, but, is it really so much confirmation that any time you think should I add more salt? you should definitely do it? If you just sit there over your baking project and pour salt in ... will that work? Probably not.
So one thing you can do is you look at the ideas you have in your brain, and each time one of them proves good, you make it a little stronger. Just a little. The tiniest bit. Like, multiplying it by 1.1. Or 1.0001. Whatever. You just give it that much extra strength in your head.
From now on, it's not a 30% chance you should add salt. It's a 30.03% chance. Or whatever. If you actually rolled it.
And maybe you add salt again in the future. And it works. So now it's like 30.06%.
And eventually you add salt and it's a mistake, so you multiply the odds by .99999% instead, or whatever.
In practice, the number stuff is really only useful if you're a computer, or you're actually randomizing your actions. And conversely no one's actually precise and unbiased enough when not randomizing their actions using a formal random number generator or whatever to multiply their chance of doing whatever by 1.0001 or .99999 or whatever otherwise. But the idea of "this worked, so I can take the parts of my thought process that were for it, and give them just a teeny tiny bit more credence in the future, not much!! Explicitly just a tiny bit!" or the converse if it didn't, extends beyond the probabilistic context.
Conscientiously applying this in every aspect of your life is known as broad-scale oopsy-doopsy epsilon-delta monte carlo conscientialism, and arguably does serve as a useful counterbalance against, e.g., the seductive allure of ethical median maximalism and its kin.
yousharknotpass
snowghoul
titsay
baratheonpilled
jbbartram-illu
