MIT Moral Machine

[aero]

Check this shit out: http://moralmachine.mit.edu/
Basically you decide what a self driving car would do in a scenario where it needed to crash due to a break failure. Post your scenarios and discuss self driving ethics.

My results:

[CynicHost]

This is a cool concept, but I find it kinda stupid because the car has no idea who the people are - and it's always better to sacrifice yourself for others, no matter who they are.

[aero]

So you're saying that if you are in a car full of babies, and would plow into a bunch of criminals by default it's better to sacrifice yourself (along with all those babies)?

[Yoshi021]

This is interesting

Spoiler: show

The only weird thing is that it tells me that I favored exercised people, but I remember that I chose both groups in a balanced way.

[YellowMe]

Oh, I wasn't really expecting this:

Spoiler: show

The part about pets is really accurate though. I wonder, why does it spell humans like that?