Thread: Self-Driving Cars and Ethics

Since we're already assuming, I will go on and assume that future self-driven cars will be secure enough for their passengers to survive the collision.

Originally Posted by GReaper

Since we're already assuming, I will go on and assume that future self-driven cars will be secure enough for their passengers to survive the collision.

Makes me think of the movie "Demolition Man," in which the cars would fill with foam on impact.

Self-driving cars are an anomaly, I think. The technology itself is well reasoned in that self-driving cars aren't really self-driving, but rely instead on humans to make almost all decisions. Self-driving cars are instead auto-pilot cars. The technology however is being presented as some form of self-driving on the false assumption that we dominate AI development. There's a gap between what the technology really offers and what is being said about it. So, it is very likely that (bar a world of fully automated roads) you won't get to witness in your lifetime those type of quandaries. To allow an on-board computer to process and resolve a collision, when not even our best aeronautical computers do it, is not going to happen until we move past the seemingly impassable barrier of weak AIs.

Originally Posted by Matticus

However, this was intended as more of a philosophical discussion, rather than a technical one.

I think the only possible background to allow an AI to reason through and act on a problem as you described, is to make the AI a first-rate citizen with the same rights and responsibilities of all other humans. Note that in many situations as you described, regardless of your choice you can still be charged criminally. And convicted. The ethical conundrum is only an ethical conundrum. It's applicability in real life is limited and most of its practical manifestations will occur under less controlled circumstances. Like the driver that needs to make a split decision about where to swerve, without any certainty as to the outcome.

For an AI to be able to reason through a similar case, there would have to be laws in place that regulated trolley problems. Laws that both humans and AIs had to obey. Without such laws (and the extreme unlikelihood of such laws ever existing in the future), a human has long learned to understand the problem of facing the consequences of his actions. For an AI to do the same it would have to be held accountable the same as humans. And for that accountability to be on par with human accountability, an AI would have to share with humans other properties, like fear, pain, love, etc. Otherwise there would be no point in sending a machine to prison for 3 years for involuntary man-slaughter.

I can't comment on what's right; there's no right answer. You haven't even considered the potential influence that such a death would cause. For example, is the pedestrian you saved a genius, hence influencing the world in a good way? Or is the driver a genius? Or whatever. These kinds of decisions just can't be made in any sane manner. I'd be surprised if there was any sort of regulation for this sort of thing in a loooong time.

So my guess is: the car breaks as hard as it possibly can such that it keeps its driver alive and healthy, and if that kills the pedestrian, so be it. Simple logic: keep the driver alive. Why? Because otherwise people would be scared of using driverless cars. I think it's most logical that people would be terrified if they think they might die in a driverless car because said car decides they should die so that someone else lives rather than the logic that someone else may die to keep you alive. Because that's just how our minds work. Put yourself before others. A basic instinct.

Hope that makes sense.

Originally Posted by Matticus

In fact, certain places are already beginning to introduce legislation pertaining to driverless vehicles. I just don't think trolley problem situations will be addressed until the technology becomes more mainstream.

Or any other type of technology, for that matter.

Let's go back a little, so you understand what I mean.

Currently the trolley problem, as a thought process as it is, is unsolved. The debate about whether the lives of 5 people are more important than the lives of one people is terribly nuanced. If we were to legislate on the matter, the complexity of the problem would present itself in full force. "Is it really lawful, just to say that for all cases, the lives of 5 people are more important than the life of one person?". The ethical debate here is endless. What if that one person is Shakespeare? I don't know... can I even measure anyone's worth in numbers of people, Shakespeare or not? In other words, isn't that in itself ethically insane?

Should I kill him before he publishes his works and save those 3 despots, 1 pedophile and that another one who is an arsehole? What if he is already an old man and can't write anymore? Or, why exactly is the life of a child more important than a woman in fertile age? Or the life of a productive man, less important than that of a lazy kid?

Conversely, if we know nothing about the people involved, how can we consubstantiate legislation when our legal systems are based almost entirely on the needed minutiae that can ensure fairness across a wide spectrum of possibilities through exceptions and addends?

So then one day we evolve our AIs. We are ready to give them the power to make autonomous decisions. We know they will always make the right decision, we just need to decide what exactly the right decision will be so we can input it. Where does this leave us?

We are back to the drawing board, trying to solve the trolley problem. A problem to which you cannot find a good decision. So, we cannot input any data of this sort it into our AIs. And while we can't decide for ourselves, we can't have them decide for themselves who lives or dies. For which reason self-driving vehicles that decide who lives or dies won't happen even if the technology presents itself.

I can't see this thought experiment becoming a real problem... at least not one that nations are going to work very hard to solve.

First, recognize that the road was already taken away from pedestrians with the advent of the earliest cars. Walking illegally in the street is known as jaywalking, and this is mostly a civil offense to protect people from accidents.

I don't see much changing to be honest. A determined person will step in front of an oncoming automatic car and even if it initiates the breaks, physics will run you over. There is no decision for the car to make, at a certain point. I imagine when it happens it will be handled much like regular car accidents. The person who owns the vehicle is responsible for what happens in/around the vehicle at all times.

EDIT: I'm glad I will die before automatic cars are dominant. I personally can't handle the transition. I'm too worried.