Think about it, are we about to put our fate on the highways in the hands of computers, computers which operate on programs based upon thousands and thousands of "what it" scenarios governing how they will respond? Or are we better off trusting man, a weak and often faulty character who nevertheless generates the scientists and engineers who design and prepare the very driverless vehicles that are claimed to be so fail safe? I think there are a lot more questions than answers and we need to be very careful before we hop on the nearby overcrowded and fast moving interstate while trusting a computer to keep us safe.
Now what made me think of this aspect of modern life? Well, I recently watched the story of US Airways Captain Chesley Sullenberger, the man who landed his plane in the Hudson River after a bird strike when taking off from LaGuardia Airport on the way to Charlotte. Most will remember this miraculous story which ended up with all one hundred fifty-five passengers and crew living to tell the story. "Sully" is a former Air Force fighter pilot with long term service as a commercial pilot, safely delivering over one million passengers to their destinations safely over four decades. And during the short two hundred and eight seconds from takeoff until forced water landing in the river, the entire career of this excellent pilot was under the gun by the National Transportation Safety Board hearings, convened to determine cause.
Now it's very important that these hearings and investigations be held, for having an aviation system that is safe and optimal is critical, but the board always seems to look first for pilot error and, of course, bases most of it's determinations upon computer simulation. Models are established which include varying scenarios and they are used to determine the viability or not of the pilot's actions. But one thing that is often left our or at least not given fair consideration is the human factor and what it contains, and this was the case with the investigation of the US Airways flight. But Sullenberger held his ground, made his argument thoroughly and convincingly, and when human simulations were conducted as he requested, using real pilots in a training simulator to carry out the actions according to the computer models, he found the flaw. The models did not include a thirty-five second period in the short flight from time of the bird strike until the Captain received clearance to try and return. And that proved to be the factor that required him to land in the Hudson, the only place open enough and smooth enough to avoid massive damage and death. The computers had not been properly set to pick that up.
So looking ahead, what is the likelihood of a similar situation with driverless cars or pilotless planes in our future? And as the traffic, on the roads and in the air continues to expand endlessly, can we be reasonably certain that the computers controlling the movement safely will accurately assess the situation where many vehicles are still operated by man? And what about the susceptibility to hacking? I think the questions are huge and the answers few, so I would hope any such effort would be extremely careful and deliberate, not rushed in a manner like the automaker Tesla is already talking about. Frankly, I hope I'm dead and buried before it ever becomes common. Just my opinion, folks, you are certainly entitled to your own.