Hacker News Reader

Comment by djyaz1200 | Hacker News Reader

Autonomous cars are going to kill people, lots of people. They are going to kill passengers, bicyclists, pedestrians etc. It has already happened and will continue. Even though autonomous cars will be (and Tesla argues already are) safer than human drivers the fact that these deaths will be at the hands of machines rather than people will outrage the public. The industry will become even more heavily regulated, and there won't be any tolerance for testing unproven autonomous systems in the wild soon.

Related... it will be interesting to see how well autonomous cars handle mechanical failures as they (and their tires) age. Good drivers really distinguish themselves in the moments where the inputs and behaviors that are expected do not occur (ex: tire blow out).

View on HN

I'm confident that autonomous cars will handle mechanical failures and other "black swan" events far better than human drivers do. Human drivers are unlikely to encounter any such incident more than once, and so are unable to learn from them; but autonomous cars can learn from all of the weird situations encountered by other cars.

But you can read about them and hear from others.

Yes, but practice is how you get good at anything.

Recovering from a skid is not something you read about. It is an instinctive response that you have to practice to do.

> Autonomous cars are going to kill people, lots of people. They are going to kill passengers, bicyclists, pedestrians etc. It has already happened and will continue.

I wouldn't be surprised if Tesla (and other autonomous driving providers) directly offer customers their own insurance and liability coverage. The auto insurance companies of the world see the writing on the wall, yet have no data to accurately model or price such a product. It might be a good thing too, since it providers a second layer of incentive to improve the underlying autonomy technology.

I would be absolutely stunned if they do. They'd be always one software bug away from bankruptcy then.

That's what stop-loss insurance is for.

> the fact that these deaths will be at the hands of machines rather than people will outrage the public

I keep hearing, and used to assume, this. But observe public reaction to e.g. the Tesla autopilot accident: measured and muted.

Counterbalancing negative reactions to machines are generational shifts in our perception of technology. I haven't seen a study on this, but from memory, what was the last new (or, for franchises, first-in-franchise) film that portrayed a man-vs-machine war? The closer we get to a population of people mostly born after the mid-1980s, the better the reaction appears to likely be.

The Matrix comes to mind as a movie that portrayed the aftermath of an apocalyptic war between man and machine.

There's quite a few recent rogue AI movies, but they typically have a few characters stopping the machine rather than a war. Avengers 2, I Robot, Eagle Eye.

Maybe. But cars seem to get a free pass with regulation, since they kill about 1.3 million people a year. We have all pretty much decided that this is acceptable, and having something traveling at 50 miles an hour a few feet from where you are walking is ok.

One has to wonder, how much longer - given that trucks are becoming the new favourite tool of mass murderers.

>Good drivers really distinguish themselves in the moments where the inputs and behaviors that are expected do not occur (ex: tire blow out).

Autonomous cars will be much better at handling these types of situations - in your example, the machine will know what's going on few milliseconds after it happens and will be able to adjust speed and direction almost instantaneously. All you need is lots of sensors.

> All you need is lots of sensors.

Which rapidly reduces the MTBF at which point you're stuck with an overpriced manual drive vehicle.

Do you also replace cars when a tire gets punctured?

When the tires cost more than the car you do.

and when the sensors start failing?

System will know that they are failing. this is not rocket science.

Actually, rockets work like this.

On that note, here's an interesting interview with the analysis lead for Mission and Fault Management on the NASA’s Space Launch System program: http://omegataupodcast.net/100-system-health-mgt/

And this is one reason they will be better to rent than own. How many times have you had a perfectly drivable car with some failing sensor? In an autonomous vehicle, such sensor failures will essentially brick the vehicle until it's fixed.

If the sensors are redundant enough, a failure or two shouldn't prevent the car from driving; it'll just drop you off and drive itself to the nearest repair place.

Plus this will make these cars safer. There are thousands of people who drive cars that should not be on the roads.

I agree that autonomous cars will be much better, but mostly not because of sensors. Experience will be the difference. Virtually no drivers actually have experience handling blowouts and know what to do during one. Even the ones who have will mostly not be very effective at handling it, because it happens so rarely, and virtually nobody actively practices it and studies and trains the best techniques.

What will make autonomous cars better is that they can take all of the input from thousands of blowouts, add "training" from actual experts, and load the resulting techniques into every car on the road, ready for execution the instant something happens.

And engineers who can recognize every possible failure and make the car react to them.

There are two things where human beings far outpace AI in competence. The first is the ability to recognize when inputs just simply don't make sense and thus the intended algorithm is going to return garbage results. A related, but important issue (particularly if you talk about networking cars together) is being able to recognize malicious inputs.

The second issue is that humans are good at higher-order planning. An example as it relates to driving is what we do when we can't tell the state of the traffic light (direct sunlight, say). Humans naturally look for ways to indirectly infer the state: the status of the traffic light in the cross direction, the current disposition of cars around the intersection.

Autonomous cars will only be better at handling these situations if the engineers are aware that these situations need to be handled and take measures to make the cars aware of them. And history suggests that is sufficiently unlikely to happen to make me fearful of the future.

You're giving us humans superhuman abilities. How many people are prepared for tire blowout? Have you ever had your wife calling you asking what is this red light on the instrument panel? Heck I myself questioned myself when ABS light turned on in a middle of my trip.. Should I continue? Pull over? Will my breaks work? When the autonomous vehicle gets that signal it will most likely play it on a safe side, pull over and call service, all automatically. Me? I just kept driving...

And yes, we can deduct light color from other clues, but we can also go through a red light because we were preoccupied with our cellphones. AV will on the other hand have access to multiple cameras, looking at the light from multiple angles, possibly with some polarizing filters that will make this non-issue.

The liability question will be an interesting one--far more so than trolley problems IMO. There aren't a lot of examples where we just accept things sold to consumers killing people when used according to directions and properly maintained because "hey, sometimes stuff happens." Drugs are probably the best example. But even that isn't a case of your drugs killing someone else.

I do think it will figured out in some future where there's true autonomy and they're proven far safer statistically. But you're right that the general public isn't going to shrug when $AUTOMAKER's car runs over a skateboarder and the company says "We are committed to safety and we'll try harder next time."

There's also the uncanny valley when no one's paying attention although they're legally supposed to be.

>But you're right that the general public isn't going to shrug when $AUTOMAKER's car runs over a skateboarder and the company says "We are committed to safety and we'll try harder next time."

Unfortunately many examples from recent years (acceleration pedals, air bags, tires) show that after few weeks everybody will forget about the issue and move on.

The general population may forget but, in the cases you cite, there were significant lawsuits in each case and settlements.

> Even though autonomous cars will be safer than human drivers the fact that these deaths will be at the hands of machines rather than people will outrage the public.

Of course. If somebody is a much-better-than-average driver, then suddenly their probability of getting killed is going to be equalized with the rest of the population and may become higher as a result.

Worse yet - due to the Dunning-Kruger effect, the average driver is confident that they are much better than average.

Reply on news.ycombinator.com