Moral machine, who should driverless cars kill?

[Deleted]

Did do the moral machine test, very easy, the car follow the trafic rules. Nobady want to ride a car who will run into the wall becuse a pedestrian stranger did try to cross the road then he did have red light (pedestrian fault)

Nobady can complaine if you the driver is in the process to drive over a pedestrian stranger who try to cross the road and have green light (the drivers fault) and the car take over and drive the car into the wall.

Of course the car shall try to minimize the damage like brake hard, call the hospital, before the impact, take the third option if possible, and maneuver away as long no third party is injured.

EDIT: This reasoning is built on that its fatall to drive into a wall, if it was "real" city traffic in low speed the car can brake and it is only a smale inconvenient to drive into a wall. Especially if the AI know that it will carch and can optimise the seat position, tensioning the seatbelt, release the airbag in the optimal time, hit the wall in the angel that minimize damage etc.

[kail]

For the sake of consistency, the car should prioritize the passengers above anyone. Any pedestrian deaths caused by the driver-less cars would be at the fault of the manufacture or someone who tampered with the vehicle. A machine has no say as to who's life is more valuable and moral to save.

I would be for emergency manual override should failure occur and if the passenger decides on crashing to save others, that would be his/her own choice.

[Deleted]

Why would the machine just not slam on the brakes instead?

Pretty much this. It's almost like everybody is unfamiliar with the concept of speed limits which are designed specifically for having enough time to react and put on breaks in congested areas.

[Forogil]

3. The car will never have enough information to truly make these decisions. It is never going to know the outcome of its actions. At best it can know how many people are in each direction it can choose. It's not going to know who survives and who dies. Pretending that information is known is what makes it so incredibly obvious that it's a thought problem and not a real one.

And additionally - the car manufacturers will not design the car to obtain the information.

[Deleted]

It should value the life of the passengers over random people at all cost.

You are as much a random person as anyone else.

OT: I'd say it should value the passangers, as other people should follow the laws of passing streets.

- - - Updated - - -

Pretty much this. It's almost like everybody is unfamiliar with the concept of speed limits which are designed specifically for having enough time to react and put on breaks in congested areas.

I'm guessing this Scenario is built on the idea of the scenario being Very Extreme.

EDIT: It is, the Study is based on the premise of the car having broken Breaks.

[Noradin]

They should follow the idea of duty of care to the passenger. Makes it much easier to model.

Yeah, occasionally that may result in bad outcomes, but no one should have to ride in a machine that they know may make a decision to intentionally kill them.

And that one is likely illegal to do because it tactily approves harming innocent and uninvolved bystanders if only to save the passenger who is the one responsible for the vehicle--and why would you do such a thing? Because they are the one who paid for the vehicle, so basically such a programming would amount to getting money for making innocent bystanders take the fall for something the aren't responsible for.
I think it is highly immoral to allow such a thing to happen.

[primalmatter]

Its my damn car... I want it to save my life at all costs.

[Noradin]

It should value the life of the passengers over random people at all cost.

It is actually the duty of the state to make such an action illegal.
The protection of random people (as in: innocent bystanders; if they were involved then they wouldn't be "random") who had no hand in bringing the situation to pass should have a higher priority than that of someone who willingly and knowlingly used the vehicle and thus is at least in part responsible for the situation.

[Daerio]

This entire "experiment" has no real application to reality, or self-driving cars.

This is basically just a public opinion poll on morality, for shits and giggles.

Under what contrived circumstances would such "decisions" need to be made by the car? Totally unrealistic.

[Deleted]

White lives should have the highest priority of saving

[Deleted]

This entire "experiment" has no real application to reality, or self-driving cars.

This is basically just a public opinion poll on morality, for shits and giggles.

Under what contrived circumstances would such "decisions" need to be made by the car? Totally unrealistic.

The brakes breaking. Kinda obvious.

[Daerio]

The brakes breaking. Kinda obvious.

Emergency brakes have existed in cars for a very long time now.

Contrived scenarios for two sets of brakes going out at the same time have not.

In fact, basically every contrived scenario you could come up with to figure out which people should preferably die, you could come up with an alternate safety feature to circumvent the entire scenario. You know, something with zero fatalities.

The entire idea of an "out of control car" is really a myth when there isn't a human at the wheel. That doesn't mean accidents won't happen; but this "moral machine" angle is total bunk.

I guess it's fun for everyone to offer their opinion on something that doesn't really matter.

[Deleted]

Emergency brakes have existed in cars for a very long time now.

Contrived scenarios for two sets of brakes going out at the same time have not.

In fact, basically every contrived scenario you could come up with to figure out which people should preferably die, you could come up with an alternate safety feature to circumvent the entire scenario. You know, something with zero fatalities.

The entire idea of an "out of control car" is really a myth when there isn't a human at the wheel. That doesn't mean accidents won't happen; but this "moral machine" angle is total bunk.

I guess it's fun for everyone to offer their opinion on something that doesn't really matter.

"The Moral Machine is a platform for gathering a human perspective on moral decisions made by machine intelligence, such as self-driving cars"

Real or not, chanse or not, it's a hypothetical experiment that is inherently hypothetical.

It's hardly for shits and giggles, if the data is parsed. You are just ignorant of the purpose of the data.

[Daerio]

"The Moral Machine is a platform for gathering a human perspective on moral decisions made by machine intelligence, such as self-driving cars"

Real or not, chanse or not, it's a hypothetical experiment that is inherently hypothetical.

It's hardly for shits and giggles, if the data is parsed. You are just ignorant of the purpose of the data.

It's hypothetical with no real application ever to be found. That's pure shits and giggles.

There's nothing scientiffic about it, you're asking for public opinions.

Just because it's parsed data doesn't mean it has significance.

[Deleted]

It's hypothetical with no real application ever to be found. That's pure shits and giggles.

There's nothing scientiffic about it, you're asking for public opinions.

It can be used for statistical representation in terms of Moral considerations, due to random selection of sampling in people participating.

Or parsing, in terms of something that would require such numbers.

You still do not know how the data is represented. You can't assert that it's for fun. That's a person belief, which you post in the same post as "It's not scientific" - Which nothing has even implied it to be.

It's a set of data that is being gathered ; You do not know what the parsing of the data is.

[Doomislav]

Driverless technology is a bad idea. Putting these types of situations in the hands of software will just juggle the liability of accidents back to the car manufacturers. So after a driverless car does something like run over a pedestrian to 'save' it's driver, the lawsuits will hammer the car manufacturers until this technology goes away.

[Daerio]

It can be used for statistical representation in terms of Moral considerations, due to random selection of sampling in people participating.

Or parsing, in terms of something that would require such numbers.

You still do not know how the data is represented. You can't assert that it's for fun. That's a person belief, which you post in the same post as "It's not scientific" - Which nothing has even implied it to be.

It's a set of data that is being gathered ; You do not know what the parsing of the data is.

I guess assigning significance to insignificant things is just part of the human condition.

[Unionoob]

This test actually opens new questions about driverless cars. If car manufacturers will start implementing such systems in cars then potential buyers will like to know know who will car save in extreme situations. And would car manufacturers even dare to answer this question or hide it.

[Deleted]

I guess assigning significance to insignificant things is just part of the human condition.

A study would hardly be done without directive ; And it's illogical that you prioritize your own ignorant standpoint, above the possibility of that you are, in fact, ignorant.

In other words ; You seem to find it easier, to shift blame on a reason, rather than admitting that you don't know what it's for.

- - - Updated - - -

Driverless technology is a bad idea. Putting these types of situations in the hands of software will just juggle the liability of accidents back to the car manufacturers. So after a driverless car does something like run over a pedestrian to 'save' it's driver, the lawsuits will hammer the car manufacturers until this technology goes away.

Or it will force the company to make cars that are inheritly hard to have malfunction.

They already pose less risk than humans driving.

[Daerio]

A study would hardly be done without directive ; And it's illogical that you prioritize your own ignorant standpoint, above the possibility of that you are, in fact, ignorant.

In other words ; You seem to find it easier, to shift blame on a reason, rather than admitting that you don't know what it's for.

- - - Updated - - -



Or it will force the company to make cars that are inheritly hard to have malfunction.

They already pose less risk than humans driving.

I could keep telling you you're wrong all day, but what would be the point?