I may be far too enthusiastic in pushing for autonomous cars, but every time I read something like this, I start to roll my eyes:
TAKING CONTROL OF AUTONOMOUS IMPLICATIONS
http://www.autonet.ca/en/2014/06/11/...utonomous-cars
Quote:
But what about the challenge of securing something that is more computer than car?
How will we protect the privacy of the vehicle's drive-route patterns and current whereabouts? Who will own that information - the owner or the automaker? If it's the latter, is it okay to sell that information?
For example, if there are two possible drive routes to the same destination, could a corporation pay the automaker to take the passenger past their place of business, in the hopes that they'll stop?
What about programming the car's software in the event of an accident? This programming is likened to a military targeting algorithm, which puts the self-driving car in a peculiar place, both legally and morally.
Protecting the car's software from attack will be imperative. Think that nightmare through: if the same software is used in all autonomous cars, and someone remotely accesses it, he or she now controls hundreds of fast-moving weapons in an urban environment.
|
But, thinking more slowly, I began to wonder how much of this is legitimately concerning.
1) Is route information really all that sensitive? I get that people don't want to feel like they have a tracker on their backs, but is having your route information logged in a computer somewhere (if this will even happen) really all that invasive?
2) Advertisers paying extra to have your car routed past their business. I cannot see this happening at all. My logic is this: How many billboards to you see along rail transit routes? Zero, because passengers in a train (or bus - or autonomous car) are not looking out the windows. They are looking at their phones or books or what-have-you's. Companies will realize very quickly that there's no sense in paying money to route someone past your business just so that you can ignore them - or worse, get annoyed at them for making your commute even longer. If any of this is even legal or possible at all.
Side note: Google was awarded a patent on a business model for allowing free transportation as part of advertising:
Source:
http://www.theblaze.com/stories/2014...r-free-taxis/#
But this is a very different deal than paying to route someone's car. Especially if you own the car and it isn't a private taxi company.
3) The moral argument: If an autonomous car knows it is about to wreck with another car, and it knows it can crash in a way less damaging to itself but in a fatal way to the other car or it can equally distribute the risk between its own occupant and the occupant of the other car, what should it do?
Whatever the right answer should be, I have a feeling that legally the car will be required to protect its own occupants at all costs, no matter what will happen to the other car. This isn't really so different than what laws are in place right now - so why is this an issue for concern?
4) A hacker hijacking an autonomous car and using it as a weapon. I don't know a whole lot about hacking and hijacking, but I do know that autonomy implies an ability to make one's own decisions. If the car realizes it isn't acting independently, that someone is controlling it via the web, It should be able to detect this and shut down, right? Again, I'm not sure how easy or possible it is to hack into an autonomous (and presumably well-secured) system, but hacking, hijacking, and terrorism strike me as stuff out of TV dramas.
Just my thoughts.