Martin Spano is the author of Artificial Intelligence in a Nutshell, a book that explores the mystified subject of artificial intelligence (AI) with simple, non-technical language. Spano’s passion for AI began after he watched 2001: A Space Odyssey, but he insists this ever-changing technology is not just the subject of sci-fi novels and movies; artificial intelligence is present in our everyday lives.Read more:Read more
When I was a little boy, our television started airing the Knight Rider series. For those who haven’t watched it, it's about a man-like artificial intelligence car that helps David Haselhoff outwit criminals. I was too young to understand that this was just science fiction, so one day I suggested to my parents that we buy K.I.T.T. (that's what the car was called). My parents didn't want to spoil my joy, so they just told me that they couldn’t buy it at the time and suggested I buy it myself when I grew up. As my wife often jokes that I still need a few years to really grow up, this prophecy will one day be perfectly true (laughs).
Autonomous cars are the dream of many. According to forecasts, their complete deployment could save more than a million lives on the road a year, not to mention several millions of injuries. The time people usually devote to driving could be devoted to more useful activities. At the same time, road traffic between autonomous cars would make road congestion much less common than at present. In the world of autonomous cars, we would not even need our own car or a driver's license as an autonomous car from the fleet of waiting cars would always come to us on request within a few minutes to take us where we need to go. How far are we from this dream?
Simply put, we are very close but at the same time very far. As we have explained in the previous sections, thanks to advances in artificial intelligence, autonomous cars have also made progress. If I had to put it in numbers, until ten years ago, autonomous cars were 30% ready; they are now somewhere around 90% ready. The problem, however, is that we did not expect the last 10% to be much more complex than the first ninety.
Autonomy versus moralityRead more:Read more
This last ten percent is made up of various borderline cases when a person has to problem-solve. Consider the following situation: An autonomous vehicle goes on a road near a cliff. Suddenly, small children appear on the road. What should the car do? Kill the children and save the passengers or turn the vehicle towards the cliff, saving the children but killing the passengers? Consider sitting in the car yourself. Would you buy a car knowing that it would intentionally kill you in an extreme case? Wouldn't you rather have a car that will save you at all costs?
A significant problem is also the unresolved legislation. Take, for example, a situation where an autonomous car kills someone by an algorithm error. Who is responsible for this? The passenger who may not even have a steering wheel to correct the mistake, the company that makes the car, or the programmers who programmed the autonomous part of the car? Before autonomous vehicles can take control of our roads, we must answer such questions.
I am convinced that we will be able to solve these problems within ten years, and the 2030s will be a time of massive deployment of autonomous cars. At least because of the promise of millions of lives saved, we should definitely strive for that.
6. Jul 2020 at 10:05 | Martin Spano