A Study on Driverless-Car Ethics • Hacker News Board

Re: A Study on Driverless-Car Ethics

#51

Post by tuacker » Thu, Jan 31, 2019, 9:41 AM UTC

cortic wrote: post #37

b_tterc_p wrote: post #5

“ if a car detects a sudden obstacle—say, a jackknifed truck—should it hit the truck and kill its own driver, or should it swerve onto a crowded sidewalk and kill pedestrians? A human driver might react randomly (if she has time to react at all), but the response of an autonomous vehicle would have to be programmed ahead of time. ” I hate this logic. No, it didn’t need to be programmed ahead of time. It should try to…

Should a car, generally prioritize the survival of its occupant or pedestrian? Not answering this is the same as answering it with ' surprise me '. And either answer is wrong.

Aside from a plane dropping out of the sky onto the direct road ahead, would a sufficiently well implemented self driving car ever end up in such a scenario?

The most basic rule of driving is to adjust your distance & speed to the surroundings/conditions. You should always be able to come to a complete stop without hitting anything. Obviously we humans suck at taking everything into account, but would the same hold true for self driving cars?

Re: A Study on Driverless-Car Ethics

#52

Post by cortic » Thu, Jan 31, 2019, 11:17 AM UTC

tuacker wrote: post #51

cortic wrote: post #37

Earlier quoted context omitted.

Should a car, generally prioritize the survival of its occupant or pedestrian? Not answering this is the same as answering it with ' surprise me '. And either answer is wrong.

Aside from a plane dropping out of the sky onto the direct road ahead, would a sufficiently well implemented self driving car ever end up in such a scenario? The most basic rule of driving is to adjust your distance & speed to the surroundings/conditions. You should always be able to come to a complete stop without hitting anything. Obviously we humans suck at taking everything into account, but would the same hold t…

Swerve (with 60% chance of survival) or break (with a 40% chance of survival). This may sound like sci-fi, but its simple math; if you know the distance of the hazard, coefficient of friction and speed. And you can crunch numbers really fast, these percentages are part of the decision process. And one puts the public at risk and the other puts the driver at risk. Now even if you had no more information and a utilitarian programmer decided to go with the numbers and swerve; swerving off-side will generally endanger the driver and near-side generally will endanger the public.

These decisions seem unavoidable.

Re: A Study on Driverless-Car Ethics

#53

Post by cortic » Thu, Jan 31, 2019, 11:19 AM UTC

rootusrootus wrote: post #38

cortic wrote: post #37

Earlier quoted context omitted.

Should a car, generally prioritize the survival of its occupant or pedestrian? Not answering this is the same as answering it with ' surprise me '. And either answer is wrong.

Neither. It should just try to stop. Calculating fatality probabilities is way out of scope.

so, 'surprise me' then, i'm sure it will.

Re: A Study on Driverless-Car Ethics

#54

Post by cortic » Thu, Jan 31, 2019, 11:24 AM UTC

mikeash wrote: post #39

cortic wrote: post #37

Earlier quoted context omitted.

Should a car, generally prioritize the survival of its occupant or pedestrian? Not answering this is the same as answering it with ' surprise me '. And either answer is wrong.

You could ask the same question about human drivers, yet zero attention is paid to this question, and zero time is spent on it during driver training.

We spend a hell of a lot of time in court deciding what was right and wrong, and when we should brake or swerve. But we have the category of 'mistake' that the program will not be able to hide behind.

Re: A Study on Driverless-Car Ethics

#55

Post by Spooky23 » Thu, Jan 31, 2019, 12:22 PM UTC

dnbgfher wrote: post #19

Spooky23 wrote: post #12

Earlier quoted context omitted.

What’s the reliability of the risk assessment? If you child died because there was a 1% higher projected chance that two people would be killed in a collision... would that seem just? I think the insistence on experimenting with autonomous driving outside of controlled access highways is insane.

And what about the family of those two people? I don't understand the insistence of bringing everything back to "what if it was your child." Pretty much everybody has family, and friends, and all that. I'm not sure how somebody being a child of somebody is supposed to be a rational argument for adjusting any of our thinking. As far as I can tell, it's basically the ultimate appeal to emotion.

It’s a call to humanity, not emotion.

When you’re talking about selecting who dies with a 1% margin with some ML process with an unknown margin of error, I question your judgement.

Re: A Study on Driverless-Car Ethics

#56

Post by maxerickson » Thu, Jan 31, 2019, 1:38 PM UTC

csours wrote: post #23

I think philosophy and culture are important, but this meme is very silly. Personifying self-driving systems is a great disservice and misleading. A self driving car detecting a sudden obstacle will not have time to classify anything to the extent that these stories have in mind, and certainly not to the point that humans would (leaving aside the fact that humans also would not have time to classify the items at a "s…

The answers to 2, 2 and 4 are pretty clearly no.

For 1 and the first 2, dynamic tolls are the smart solution. Also applies to the second 2. If people want to go faster, let's tax them for it.

Re: A Study on Driverless-Car Ethics

#57

Post by tsukikage » Thu, Jan 31, 2019, 2:44 PM UTC

RcouF1uZ4gsC wrote: post #25

b_tterc_p wrote: post #5

“ if a car detects a sudden obstacle—say, a jackknifed truck—should it hit the truck and kill its own driver, or should it swerve onto a crowded sidewalk and kill pedestrians? A human driver might react randomly (if she has time to react at all), but the response of an autonomous vehicle would have to be programmed ahead of time. ” I hate this logic. No, it didn’t need to be programmed ahead of time. It should try to…

I agree with you that this example is so theoretical as to be useless. In addition, focusing on this masks some other more relevant ethical/legal questions. 1. If the car sees a police car flash it's lights for it to pull over, should it pull over unconditionally - not allowing the driver to over-ride? 2. If there is an Amber alert broadcast specifying the car, should it automatically try to alert the authorities as…

We have the capability right now to make cars that enforce speed limits on their drivers.

We do, in fact, make cars with audible / visible / tactile alerts when the driver exceeds the speed limit for the road they are on, but we seem to have no appetite for buying cars that take the final step to actually automatically brake, never mind requiring all cars to do this.

I take this as a very strong indication that overriding driver preferences as in 1-3 will not happen any time soon.

Re: A Study on Driverless-Car Ethics

#58

Post by andrepd » Thu, Jan 31, 2019, 3:04 PM UTC

cortic wrote: post #52

tuacker wrote: post #51

Earlier quoted context omitted.

Aside from a plane dropping out of the sky onto the direct road ahead, would a sufficiently well implemented self driving car ever end up in such a scenario? The most basic rule of driving is to adjust your distance & speed to the surroundings/conditions. You should always be able to come to a complete stop without hitting anything. Obviously we humans suck at taking everything into account, but would the same hold t…

Swerve (with 60% chance of survival) or break (with a 40% chance of survival). This may sound like sci-fi, but its simple math; if you know the distance of the hazard, coefficient of friction and speed. And you can crunch numbers really fast, these percentages are part of the decision process. And one puts the public at risk and the other puts the driver at risk. Now even if you had no more information and a utilitar…

These conundrums are irrelevant anyway. Over one million people die every year on the roads. The instant AI becomes better than humans, it's a moral imperative to adopt it, no matter what it does in these rare and contrived circumstances.

Re: A Study on Driverless-Car Ethics

#59

Post by rootusrootus » Thu, Jan 31, 2019, 3:17 PM UTC

seventhtiger wrote: post #46

rootusrootus wrote: post #38

Earlier quoted context omitted.

Neither. It should just try to stop. Calculating fatality probabilities is way out of scope.

Years down the line, a brand of self driving cars will market themselves as the safest care you can buy, exactly because they prioritize the occupants of the car. At some point it will have to be regulated.

That would be saying that it will aim for pedestrians. If someone designs a car with that intent, then yeah, it will either be regulated or sued out of existence.

Re: A Study on Driverless-Car Ethics

#60

Post by rootusrootus » Thu, Jan 31, 2019, 3:25 PM UTC

cortic wrote: post #53

rootusrootus wrote: post #38

Earlier quoted context omitted.

Neither. It should just try to stop. Calculating fatality probabilities is way out of scope.

so, ' surprise me ' then, i'm sure it will.

No, I said 'Neither.' You presented a false dichotomy.