Intelligence is not Artificial
by piero scaruffi
Cognitive Science and Artificial Intelligence
| Bibliography and book reviews
| My book on consciousness
Back to Cognitive Science
| My book on consciousness
| My reviews of books
(These are excerpts from my book "Intelligence is not Artificial")
The Dystopia of Vast Algorithmic Bureaucracies - Appendix: the Self-driving car
Self-driving cars are a mirage in real-world traffic. But the vast algorithmic bureaucracy could simply introduce rules and regulations to turn real-world traffic into highly structured traffic, i.e. traffic where anybody could drive a car, even someone who never took driving lessons.
In my opinion, adoption of the self-driving car will be driven not by Artificial Intelligence but instead by economic interests that eventually will impose new limitations on our lives in order to make the self-driving car a reality. As of May 2017, Uber boasted one billion passengers and 1.5 million drivers worldwide. More than 50% of the revenues were pocketed by the drivers. Hence Uber's motivation to get rid of the drivers. In 2015 Uber hired more than 40 scientists from Carnegie Mellon University's prestigious robotics laboratory. This is the laboratory led by Red Whittaker that won the first DARPA Grand Challenge for autonomous vehicles in 2004 in the Mojave desert and that came second in 2005 to a Stanford vehicle built by their former scientist Sebastian Thrun, and then won again in 2007 DARPA's Urban Challenge. Red Whittaker had been building robots since 1979 when Carnegie Mellon University helped clean up the mess of the nearby Three Mile Island nuclear reactor accident. Unfortunately the self-driving car doesn't work, and it won't work any time soon, so firms like Uber will inevitably come up with a scheme to pressure cities to limit what people can do. If we turn streets into railways with no rails, then the self-driving car becomes a reality.
The big car manufacturers would be perfectly happy with this outcome. Ride-sharing firms like Uber don't make cars and someone will have the make Uber's self-driving cars. The advantage is that firms like Uber will standardize what a self-driving car will look like and its features. If the cars of the future are purchased by Uber's standardizing bureaucrats instead than by 100 million idiosyncratic consumers, the car manufacturers will be able to save a lot of money and earn a lot more of money.
Last but least, car manufacturers may have smelled another source of revenues: providing remote driving assistance. A startup called Phantom provides a remote driver who can see what the "self-driving car" is doing and can take over the wheel remotely via a fast wireless telephone line.
Economic interest is, to me, the real driver of the self-driving car (sorry for the pun).
"Knowing what I know about computer vision, I wouldn't take my hands off the steering wheel" (computer vision pioneer Jitendra Malik in 2016)
Back to the Table of Contents
Purchase "Intelligence is not Artificial")