When an autonomous car has to make a decision, will self-interest or the public good predominate?
Originally shared by Next Nature Network
When an autonomous car has to make a decision, will self-interest or the public good predominate?
http://buff.ly/2bBOAOb
When an autonomous car has to make a decision, will self-interest or the public good predominate?
http://buff.ly/2bBOAOb
While I acknowledge these questions are valid, they are the equivalent of asking everyone what they plan to do with a $300million lottery payout. Yes, for some incredibly small percentage of humans on the planet, it will make sense to have planned for such an occurrence. But for 99.99999999% or so of us, developing such a plan is simply fruitless daydreaming.
ReplyDeleteThe lack of distractions, the lack of fatigue, the constant learning, and the communication with other vehicles and monitoring stations will decrease the chances of the vehicle having to make such a choice from nearly none at the introduction to not-worth-mentioning after a year of significant market penetration.
OTOH, as these robots and others take our mundane and meaningless jobs from us, we'll need something to occupy our minds I suppose.
As an engineer you spend 90% of your time on the not-so-obvious corner cases. Also, as part of the team working on one of these (not myself, but just imagining) you would have to consider every accident at least in part your fault. Even if that's not fair, psychologically it's there.
ReplyDeleteIn that way, you can basically say that the team of hundreds of people or so (relatively few) will be driving the hundreds of thousands of cars that are driver by their engineering.
So to me this is exactly the type of thing that needs thought about.
I would never say that the questions are not needed. However, they have received a prominence in the minds of the general public that is undeserved (inasmuch as the general public knows and/or thinks about such issues). To me it looks more like lawyers laying the early ground work for instilling FUD into potential jurors for the inevitable lawsuit when something no easily explained happens.
ReplyDeleteDriving simulators of years past, video and telemetry from current autonomous vehicle programs, and real world experience will feed AI systems so that there will in fact be less culpability by the engineers. That is why all of the cars will be better drivers than any of the the humans. But no, perfection is unlikely at least initially. However, they are already safer drivers than the vast majority of humans. As a bicyclist and motorcyclist(and engineer) I would much rather share the road with current technology autonomous vehicles than the mass of people I see driving while talking on the phone, texting, eating, adjusting the radio, putting on makeup, checking on the kids in the backseat, trying to get a better look at the hot chick in the next car over, etc etc etc.
I agree. The FUD is inevitable. So are the lawyers. If you want to slow down a tech the easiest way is FUD.
ReplyDeleteThe only way out of the FUD is to prove that it's well thought out and that the feared situations are thought out. After that initial hump of surpassing the FUD is over complacency will take over and it'll be smooth sailing.
The same thing happened with cars on the roads. The auto industry played into it and used it in that case. (See http://www.trutv.com/shows/adam-ruins-everything/blog/adams-sources/adam-ruins-cars.html ) I don't see that working so well in this case. But once the companies like Google and others are ready to pull the trigger, the FUD will get cleared by marketing real quick, I'm sure.