Autonomous driving simulators may help reduce driverless anxiety


#1

Originally published at: http://boingboing.net/2016/12/09/autonomous-driving-simulators.html


#2

Much of my anxiety is caused by knowing that human drivers are already autonomous - yet nobody tries actually linking them together to drive as a network.


#3

They also know the future is autonomous driving

Are they also selling a device for knowing the future? I want that.


#4

I’ve used the device to look into the future, and in all of the futures where you got the device… Well, let’s just say you don’t really want it.


#5

Can we get an automated system for the poor gamers playing Desert Bus? :stuck_out_tongue:


#6

I could watch dash cams on youtube all day long, it still won’t reduce my anxiety about driverless cars, because my anxiety is (at least partly) rational. It’s not that looking out the windshield gives me the willies - I can have that experience any day of the week, it’s called the passenger seat. No, my anxiety is because automated systems are not reliable enough to make me want to put my life in their …hands? Mandibles? Whatever.


#7

Out of curiosity, what would constitute adequate evidence that a given autonomous system is sufficiently reliable? Or at least, as reliable as systems you already trust with your life? As @popobawa4u points out, other humans are autonomous and unreliable, and I’d be willing to bet you get in cars while some of them drive.

Similarly, what evidence would be enough to convince you that all the other drivers on the road should be replaced with autonomous vehicles?

I would observe that if there is no possible evidence that would convince you, then the problem is not with the reliability of the autonomous system.

I wouldn’t trust a fully autonomous car today, nor the word of any company or engineer that assures me some new system is safe. But at some point someone could have a million cars on the road, with tens of billions of total miles driven, in all weather conditions, running the same software. Suppose they then publish how many fatal accidents, how many injuries, and how much property damage they have caused or been subject to in collisions. Suppose you and anyone else are permitted to audit the hardware and software for security flaws. What then?


#8

You’ve been around long enough to know better. A track record of success would do the job, along with falling death rates. But I’m not an early adopter. Frankly, I think it’s much too early to talk about autonomous cars, the number of hurdles is still large.


#9

Then I agree with you. I am not generally an early adopter either.


#10

I have no issues with an autonomous vehicle taking me around town or the country. I have issues with the other drivers on the road that may freak out and try to override the car to take control. Those are the things that bug me.


#11

Well, my point was not that humans are unreliable, but that billions of dollars are not being put into networking the cars when no such resources were put towards networking and coordinating the autonomous drivers we already had.


#12

I see. Pretty sure that’s what road rules and signs are meant to do,
though. It’s a less formal system only because human minds are bad at high
bandwidth real time many to many communication. And we’ve definitely spent
billions on developing those over the years. Compare to countries that
don’t have road rules enforced much if ever.


#13

See, I’m fine with the general idea, too. My issue would be that my autonomous car is not “making good time” compared to the others. Just let me pass, and we’d be cool.


#14

THAT to me is the far more interesting problem to throw such funding at - instead of moving billions of people around so they can buy, sell, or manufacture pop-tarts. Solve problems in a way that increases that capabilities of the organism itself, rather than diminishing it.


#15

I would like to see that too. I think it is a much harder problem though.


#16

This topic was automatically closed after 5 days. New replies are no longer allowed.