The Fifth Generation by Pamela McCorduck and Edward A. Feigenbaum published 1983 delves deep into the problems of artificial intelligence (AI), and the perceived dominance of Japan in the race for the supercomputer that mimics human intelligence. McCorduck had already published a book, Machines Who Think: A Personal Inquiry into the History and Prospects of Artificial Intelligence, an inquiry into that subject first published in 1979, and amazing for a computer book, republished twenty five years later in 2004. These books were direct, and serious inquiries into problems which were at the time well over into the realms of science fiction. But, these were not speculative fantasy; these were careful conversations with what turned out to be many of the founding fathers of the computer revolution.
From the beginning there has been the worry that the machines which humans have created to ease their human physical labors, and mental exertions might in the end replace their human masters entirely. Even Homer’s Iliad had worrisome robots. To those ancient people what we, having had more exposure, would see as simple machines were seen as mysterious, and were used by their unscrupulous leaders to awe them into obedience. Then along came the Golem, and Frankenstein’s monster, HAL, Deep Blue and now Google, Second Life and Wikipedia. All of these non-human creations of humans possessing powers totally beyond the power and understanding of ordinary human beings. Each of these monsters has brought with its creation a kind of terror to the human community into which it inserted itself. Each of these possessed a kind of uncanny super human wisdom which grew by itself out of the simple mechanical thing which its creators had somehow animated into something much greater that the sum of its parts. All of the earlier beings had been created to serve humans, but somehow managed to escape from their creators, and wreak havoc upon humankind. The more recent ones are seemingly under human control, but for how long. Back when I was a airplane pilot I used to say to my passengers, “If you don’t like the way I fly, get out and walk.” It was a standard joke when flying, and usually got a lame laugh, but I would say the joke before I even started the engine. It seemed more fair that way. This whole computer AI thing is like that, and it seems like the engine has been started, and we are rolling down the runway. Not really fast just yet, but even now to jump out would be painful.
In some ways the computer revolution seems to be moving ahead rather slowly — after all it takes almost two years to double the power of the CPU, and the RAM, and the operating system, and the hard drive, and the numerous peripheral drives, and the various connecting speeds, and the secondary add ons like cameras, audio recorders, video cameras etc. which also have been growing apace. Of course things seem a bit faster than this slow pace because it all works together, and connects to the Internet which is also doubling at this paltry pace. The synergistic interaction makes it hard to understand what is happening, and then of course there are those overexcited kids in their garages that are trying to do something outlandish. Who knows what’s next? Movie tattoos? Remotely controlled dancing viruses? Remotely controlled people?
There is one tiny little thing about this doubling – it is exponential, and gives the impression that it is going to remain exponential for several more doublings. Furthermore, that doubling is already upon a tremendous base of accumulated skill and power. The only immediate limit in sight is the total population of humans to be accessed by the Internet. Today there is 21.1% Internet usage penetration world wide with a 290% growth rate from 2000-2008. In a couple of years almost everyone will have Internet access. When that happens anyone can know almost anything that any other person knows. As the unknowns become discovered there will be concerted efforts to find answers to those questions rather like the automatic filling in of Wikipedia entries. This will be more like questions of wisdom, and experience than questions of recorded facts and there will be real people behind the answers because when there are 6.7 billion people online there will be someone discussing almost everything conceivable live. That is already happening, but it will become more exact as the methods for finding each other improve. Also, anyone will be able to purchase almost anything they need, and have it delivered quickly in part because potential things for sale will be tracked not just offered items. To some extent this already happens when people list their personal libraries online, these are not exactly for sale, but they could be if an offer was made.
None of this seems particularly inclined towards precipitating a Doomsday, and it all seems very positive for human growth. But then there is the madness of crowds even Internet crowds where they all start behaving in some strange way. Or perhaps someone does create an unstoppable virus and it fries everything or there is a relatively minor atomic war, and the EMPs fry all of the electronics in the world. Oh blather its time to worry about something else.