The last two times I've been in a fight went exactly the same way.

Dumb drunk guy swung at my face, I took it, a bunch of bystanders jumped on him and hauled him off. Pretty much end of story.

I've plenty of fighting experience, the ones that have ended badly for me have been the ones where I've fought back.

Obviously not the trolley problem-esque situation from the context, but my core point is that one cannot construct morality from extreme hypotheticals.

▲

9x39 a day ago | parent [-]

> my core point is that one cannot construct morality from extreme hypotheticals.

Isn’t that what happens when we codify limits of behavior, which are often extreme, into laws or religious texts which then govern a population?

Even if you don’t consider law as de facto defining morality, moral lessons from literature to oral tradition are often handed down as metaphor through stories of finding balance between extreme outcomes.

	▲	specproc a day ago \| parent [-]
		Yes, we take extreme behaviour, that is harmful to others, and prohibit it as a society. Where we run into trouble is where we say, well here is an example of a case where this extreme behaviour may be countenanced. There may well be such cases! If I had a gun (I don't) and someone was attacking a loved one (fortunately rarely if ever happened) with intent and ability to kill (definitely never) and the only way I could stop them because of the specific situation was to kill them (waaay down the tail now)... perhaps, I don't know how I would react in that situation. Here's where ethics becomes like programming. I could sit down and come up with a list of cases in which I felt it appropriate to kill, and code all the edge cases. This is inefficient and sorta silly, and I guess how I chose to interpret the comment that started this thread. I could come also up with a clever algorithm which balances harm done and harm prevented (or good caused) based on a range of parameters. I think this is more what GP was pointing to, a teleological ethics. But what model? What parameters? What loss function? Which libraries? My position here, at least on violence is deontological. If everyone can write their own crufty (and inevitably closed-source) solution to the problem, then bad actors can (and do) code it so they get the results they want when they need them. The result is a violent world. The cleaner, more elegant, and more ethical code simply prohibits harmful outcomes altogether. I suspect derivations from this initial simplicity in religious texts and interpretations are malicious code added in later updates.