[dx034]

It's not that unrealistic. Imagine driving a car and a kid runs onto the street. You can swerve onto the sidewalk but there are people walking there you'd hit. The question if you'd take action to save someone to (potentially) kill someone else is interesting. And it has been discussed in the context of self driving vehicles.

But from all I remember, the outcome always was that an AI should never actively make a decision to sacrifice someone. And that's also how I view the trolley problem. Actively making a decision feels worse for me, even if fewer people die in the end.

[Sirened]

> You can swerve onto the sidewalk but

I feel like that's not realistic either—when you encounter these sorts of situations in real life, you don't really make a thought out choice because it happens so fast. People swerve and crash into trees trying to avoid rabbits, it's not a reasoned thing.

These problems do, however, come up fairly often in so many other areas. Cryptography is probably one near and dear to many people's hearts here—supporting cryptography directly saves many lives (journalists in totalitarian regimes, people in abusive relationships, the general wellbeing of people being able to communicate securely for a myriad of industrial purposes, etc.) but it also, to a lesser degree, directly leads to some deaths (terrorists are able to organize and hide their plans). Do you spend your life developing more powerful cryptographic algorithms knowing that it will have some small negative outcome that your work is partially responsible for? Or do you do nothing at all and have a larger number of people suffer as a result of you not having produced a work.

[codefreeordie]

In real life, though, you never have either perfect information about whether your action (or inaction) will have your intended consequences nor do you have the time for complete/rational analysis.

[exodust]

> It's not that unrealistic.

It's unrealistic because the trolley problem is basically a torture apparatus that you're instructed to operate for unknown reasons. Our decision to pull levers or not, means nothing in the context of why the situation exists in the first place, and why you have the job of executioner.

We have time to think and weigh things up in regards to pulling the lever. But in the driving scenario, a split-second decision is needed. Muscle memory and reflexes come into play. Everyone will apply the brakes anyway. There is no "do nothing" option in the driving scenario.

[mrwh]

Agreed re actively making a decision: my answer was mostly not to pull the lever, and if I analyse my feelings it's because then the results aren't my fault. Pulling the lever makes me party to the situation.

Disagree re realism though. You never get a chance to think in real life; and any situation in which you could deliberate what the moral course of action is, is a situation you could avoid entirely. Really the closest to trolly problems in real life are public policy decisions - which are real and affect all of us. So yes I guess I've argued myself around to agreeing again :).