Should driverless cars be programmed to kill you?

I found this interesting and thought it might spur some interesting discussion.

Current reports suggest 10 million driverless cars could be on the road by 2020. The issue is this: In a few very rare scenarios, a driverless car may have to make a choice between protecting its occupants and protecting pedestrians. For example, if it's driving down a road at speed and someone runs out into the road, should it swerve into other traffic to avoid them, potentially injuring or killing the driver and passengers? Or should it make every attempt to stop, even though it knows it won’t be able to, killing the pedestrian?


http://www.iflscience.com/technology/the-fatal-moral-dilemma-posed-by-driverless-cars/

Give me a 69 all metal no computer Mustang any day... I'll pass on the smart cars... phones didn't take the smart word too well ... let's see how cars like the title. Lol I'll wait this one out even if I have to ride the trolley! Jk

I don't like the idea of self driving cars. I'm a bit of a control freak and I don't like the idea of a program being responsible for my transportation and my safety. My phone/computer/playstation/all other pieces of technology that I own have problems way too often for me to trust them with my life. Plus, I genuinely love driving. It's so much fun. There can be self driving cars, I just don't want to be in one.

As for the morality issue...that's tough. I suppose that the people in the car would usually have a better chance of surviving the resulting crash than the pedestrians have of surviving being hit, all situation based of course. So it might be "better" to have the car attempt to miss pedestrians. That might change depending on speed and surroundings. Not an easy question.

Hehe, nah. I'd worry we'd have idiots walking out into traffic knowing they would be ok, not caring about the impact it has on others!!!

Plus I think it would be a bad idea to cross into oncoming traffic, while obviously dramatically increasing the collision force there is also that approaching vehicles would be coming from further away and therefore less likely to be informing whatever sensors the car has about the size of the oncoming traffic. Bumping into an oncoming car at slow speed might be better for everyone under ideal conditions, but what if the occupant in the other car is already carrying an unrelated injury or what if the oncoming car is a petrol tanker. They are not doing anything 'wrong' and so I think its unfair to include them in the accident by choice.... unless the vehicle suffers a mechanical fault of its own, as then it might have better odds for everyone hitting another vehicle then an unprotected person or crowd.

So to you Q, the pedestrian is doing the wrong thing, so I think the car should aim for wherever the biggest gap is, brake, slide whatever, and if it is faced with going one way into oncoming traffic or the other way into the pedestrian - I'd program it the later because the pedestrian should not be there and perhaps the sensor is incorrect in its sensing (is it really a person!?). BUT would smart cars be talking to each other, in which case we'd have a greater margin for safety by all trying to make the biggest gap for the incident.

Kyrin Wyldstar wrote: I found this interesting and thought it might spur some interesting discussion.

Current reports suggest 10 million driverless cars could be on the road by 2020. The issue is this: In a few very rare scenarios, a driverless car may have to make a choice between protecting its occupants and protecting pedestrians. For example, if it's driving down a road at speed and someone runs out into the road, should it swerve into other traffic to avoid them, potentially injuring or killing the driver and passengers? Or should it make every attempt to stop, even though it knows it won’t be able to, killing the pedestrian?


http://www.iflscience.com/technology/the-fatal-moral-dilemma-posed-by-driverless-cars/

I don't like the idea that technology is going to be making moral decisions for us, and I like driving, so I won't be owning a driverless car until the day the government forces me to. A car can't tell the difference between people, e.g. a mother and baby versus an elderly person. As horrible as it may be, in a split second my brain is probably going to go mother and baby. A car can't make that kind of decision.

Goken wrote: I don't like the idea of self driving cars. I'm a bit of a control freak and I don't like the idea of a program being responsible for my transportation and my safety.

It's the next step to getting hover cars :lol:

Can you imagine the carnage with manually flown flying cars OMG :side:

That is a situation Ive always disliked. Tech messes up far far too much for me to trust it with my life or the lives of others. Imagine a chemical tankers sensor messing up and causing it to crash to save the "most" amount of lives but there was no actual danger to anything and it crashes leaking toxic chemicals for nothing. Besides I will not put my life into the hands of something that holds no ability for compassion or reason. In short Ill walk everywhere before I trust a machine that might kill me because of a pre-programmed algorithm. Frankly I would hold the programmers of the cars guilty for every death the car causes and hold no issue with sending them to jail for it.

Greater responsibility no accountability... smh I don't think we are ready for it... human beings. We still can't decide whether or not where to pee.. my own opinion though... we focus on way to many.. frolicking like ideas some times. The road and behind the wheel... a place to relax and let us do the driving... hmm....still thinking...

Further issue. These cars would all have to be on some kind of network, I'm assuming. It seems like the kind of thing where they could be accessed remotely. What if some crazy hacker decided to mess with the programming? Heck, the idea of someone remotely controlling my vehicle at all is too much for me. They're already out there. Cars with stuff like OnStar where they can remotely shut the car down if it's reported stolen.

:laugh:

I just re-read my last post and now I feel like some crazy conspiracy theorist. I'm not saying I'm wrong, I just think I sound a bit nuts! :laugh:

Adder wrote:

Goken wrote: I don't like the idea of self driving cars. I'm a bit of a control freak and I don't like the idea of a program being responsible for my transportation and my safety.

It's the next step to getting hover cars :lol:

Can you imagine the carnage with manually flown flying cars OMG :side:

If driverless cars will get us closer to hover cars, then count me in. I always have been an off-road kinda guy.

The piece missing from the equation here is the role that the actual roads could also play in this. Driverless cars don't have to be unleashed on every side street, alley, backroad, turnpike and freeway. I imagine that sensors could be installed over highways or lanes meant specifically for driverless cars similar to the ones used on toll roads now that read transmitters present in the vehicle. The vehicle can be programmed to only allow driverless operation when on certain roads or in certain lanes where these transmitters are present, but still be manually operated everywhere else. If we keep the driverless operation to roads with little or no pedestrian traffic it would greatly mitigate the risk of these moral dilemmas.

Since these cars sense each other and set proper speeds and distances based on that information, we'd basically be turning the 'driverless car lane' into a train made of individual vehicles.

I see my morning commute as manually driving my car onto the interstate, entering the proper lane, then putting it on 'autopilot' and reading a book or drinking my tea. When it's time to get off the interstate, I take over again. It could be a glorified version of cruise control.

It would be especially helpful if it tied into the GPS navigation and warned you when your exit was coming so I could wake up from a nap and put it back into manual mode in a timely manner. :laugh:

Sorry for the derail, but on flying cars I found an electric helo with good engine redundancy, simple controls lets driver focus on nav and traffic avoidance;

https://www.youtube.com/watch?v=OazFiIhwAEs

If they could fold that rotor assembly they could be parked on rooftops next to each other under a landing/launch platform, for commercial buildings.
:lol:

While you can argue the fallacies of tech....I have seen more than enough fallacy in humans driving. Drunk driving, road rage, texting while driving, simply not paying attention, ,people who are way to old to have a license,etc,etc.

I can't imagine it would be much worse actually.

I always wonder if people simply don't like change much when these things come about.

http://www.youtube.com/watch?v=XU3qfDtWFmk

I have seen what the Tesla cars do with such beginnings of technology a d I have to say, I see little to be so negative about.

http://www.youtube.com/watch?v=2iwfNs8j2FI

i didn't read all the posts...but driverless cars come with a major risk...hackers can access them....it was already proven with a new jeep grand cherokee, a hacker accessed the jeep's computer through it's onstar thingy or whatever chrysler uses, and the hacker was able to control the radio/ac/electronics and was even able to shut the vehicle down while it was being driven...the driver had no control over the vehicle once it was shut down...

just something to think about...

edit: this was done with knowledge as a means of proving that it was possible...the driver and hacker were demonstrating the vulnerability "smart" vehicles have...

to answer the question posed by OP

imo, the car should be programmed to protect the driver - it should be a personal safety vehicle foremost (a whole new angle on the idea of "product loyalty" lol; in future, the products are loyal to YOU!), especially in a circumstance where an outside agent was violating procedure in a way that put lives at risk

as for self driving cars in general

i do not want to own a vehicle if the control of that vehicle can be forcibly taken away from me

for one thing, i imagine a good way to end the revolution would be to simply lock suspicious people in their own cars and remote-drive them to the internment camp

other than the issue of control being taken against our own will or interests, i think theyre a great idea

once the parameters have been set, computers are more capable and less fallible at acting within those parameters than humans are

LOL I have to say this went in some interesting directions! While the question I proposed was designed to be a moral one many went off in a terminator/mr-robot/blade runner direction. Thanks for the responses.

As for the question I would tend to agree with what seems to be the consensus here. If a driverless car is faced with a moral decision such as this it should protect the driver first and any outside obstacles secondary. So that means if a pedestrian stepped in front of a car and the car could not avoid the collision it should not be programmed to swerve into traffic or off a cliff etc to protect the pedestrian.

I think in order to clearly assess this situation we need to put ourselves in the role of not only the driver but the pedestrian as well. If you made a mistake and stepped in front of a moving vehicle would you want that vehicle to be programmed to swerve and kill its occupants to protect you? Now I'm sure some would say YES! But objectively speaking if you made that mistake you should take responsibility for that mistake and accept the consequences vs letting the car occupants pay for it.

So in the end I don't think cars should be programmed to swerve to avoid pedestrians or other such potential situations, (say a pregnant woman etc) even if there is a likely hood that more could be killed outside the car vs inside.

I think ultimately, the car will make the best possible choice based on the variables, and given its ability to think not only objectively, but faster than a human, it will be more than likely better than a persons choice with slower reflexes and OODA loop.

It is more than likely the person in the car will be safer regardless of choice, given airbags, seat belts, and whatever technology comes up with in the meantime.

Something like that foam from "Demolition Man", which had the option of self drive and not actually.

It will decide whether to swerve, or not, or what have you, based on the a lot of high level math in a few seconds to make the best possible outcome.

You see, the car will not make a "moral decision" at all.

This clickbait-headline-as-thread-title makes it sound more like the question is about whether driverless cars will be programmed to immediately kill their inhabitant, as if we live in some bizarre dystopia where that is the sensible solution for population control. I realize it's taken from the linked clickbait article, but c'mon, surely you could have done better than that, given the question is really far more nuanced than that.