Killer Robots and Driverless Cars
Imagine driving down the road on your way to work and being faced with a split second decision ...
Image Source pixabay.com
Little Johnny just ran out in front of your car and your going too fast to stop, would you swerve to avoid him or run him over?
...
I have dreaded the rise of the machines since I first saw Terminator 2, but August 29 1997 came and went. The doomsday scenario portrayed in James Cameron's masterpiece may never come to pass but we regularly put our lives in the hands of machines and some day soon they will decide who lives and who dies...
The Future is Now
After a recent visit to Japan between; the futuristic toilets, interacting with the loveable Pepper and trying out a Virtual Reality Experience in Shinjuku's VR Zone, the future has never seemed so close.
We are still a long way from when machines will start questioning the need for humans existence and raging war to wipe out "The Resistance", but we are already very close to having to deal with some tough ethical dilemmas.
Killer Robots and Driverless Cars
Today at work we were discussing how driverless cars were going to impact the Insurance industry. There was discussion about claims costs, liability and all the other boring stuff,
but then someone raised an interesting point:
There is a foreseeable scenario where the AI of the driverless cars will have to choose between killing the driver of the car, or another person. Not all situations are as clear cut as the choice faced when Johnny runs out in front of your car.
Consider a car driving down a road when someone steps in front of the car. What if the on board computer of a driverless car has to choose; run the person over or divert the car into a tree at the side of the road likely killing the driver ... there are countless variations of these killer scenarios, not all that far fetched!
Image Source pixabay.com
Moral Compass
Artificial Intelligence conjures up futuristic images of killer robots rising up against their human creators but the technology of today could be just as lethal.
How would you program a computer to to kill someone?
Image Source: pixabay.com
Currently computers aren't that smart or self aware. The outcome of current AI is largely based on logic or decision trees pre programmed by a human programmer. At work today I asked our resident programmer how he would go about addressing this ethical dilemma faced when a driverless car had to choose between killing person A or swerving and killing person B
...
Self Termination
The choice was clear, in his mind at least, the car would choose to self terminate. This means swerving and killing the driver thus maintaining the external environment. This course of action is a corollary Asimov's Laws better known as the three laws of Robotics where the Robot would choose not to do harm to others, and instead choose the course of action that would lead to its own destruction.
Killer Cars
A driverless car pre programmed like this would represent a car programmed to self destruct and kill its human owner (in one scenario at least).
Image Source: pixabay.com
It's now 2017 and after my visit to Japan driverless cars don't seem that far off. The science fiction future of Johnny Cabs has never felt closer.
Would sit into a driverless car knowing it has been programmed to kill you?
Thank you for reading this. I write on Steemit about Blockchain, Cryptocurrency, Travel and lots of random topics.
The more computer systems a device has, the better it can be hacked. Those cars will be hacked. For what kind of things will they be used? (Mindmap?)
(I know, we can't really vote up)
l know there are cars that can drive itself here in japan and l think its crazy! yet the cat has own safety breaks are awesome! that can prevent from my many accident 😃
The safety features on modern cars mean they are safer than ever. A computer can see and react in ways not possible by a human but, at the moment at least, the responses need to be pre programmed. It raises an interesting ethical dilemma and question of liability for car manufacturers.
Superb writing as always. Resteemed.
Thanks @condra I appreciate it.
Wait hope day introduced into the Japanese ♫
I think the future will depend on the robots will greatly we will dispense with the labor force
So scary I hope it doesn't happen to soon such a liability for peoples lives
Hmm. I think I'll stick to me bike!
Just stay out of the way of those driverless cars 😀
I've had a similar convo with friends years ago when self driving cars were a real hot topic. We both agreed we'd rather be at the wheel in said situation & that leaving AI the choice over who lives or dies in said scenario is kinda crazy!
BTW How awesome was Japan Brotha!!!
Been there myself (one of my favs)
Where did you end up, mostly Tokyo or did you have a chance to travel to Osaka/Kyoto etc?
Tokyo is another planet, just amazing. We were in Japan for two weeks in total but we packed a lot in. Kyoto, Osaka, Kanazawa, Yamanouchi and Hakone. It was a great trip. We even got to a steemit meetup in Tokyo 😁
Where did you get to when you visited?
Asimov's Laws are completely impractical to talk about because they are so ambiguous. As for talking about whether I would get into a car that has "been programmed to kill me" it is only in very specific scenarios that that would happen and in many scenarios it is very easy to get out without anyone dying. So ultimately it is impractical to even think about a situation where the car would have to kill you.
Very good questions to ask but I don't think that they will be answered for us. It's going to be expected for us to take a leap of faith, a "trust us" scenario: something of which I have a very short supply.