GDC 2008: Ray Kurzweil: The Magic ManGDC 2008 - RSS 2.0
Kurzweil says what he does isn't all that hard, but that most people just can't see the exponential curve of technological progress. He says most people view progress linearly, expecting greater advancement more quickly than it actually happens, and being surprised by the rapid explosion of advances as the exponential curve rockets upward.
He used the bursting of the technology bubble in the year 2000 as an example. Kurzweil said this was a Wall Street problem, not a Main Street problem. The explosion of popularity of the world wide web in the late 1990s led Wall Street analysts to (correctly) assume the web would revolutionize commerce in all areas, but it didn't happen as quickly as they wanted it to. They were thinking linearly. But exponential growth takes a long time to ramp up; then, when it does, it moves a lightning speed.
By 2000, Wall Street had lost patience with the web and the tech stocks crashed. A few years after, as the exponential curve predicted, the technology had advanced to the point where many of The Street's predictions came true, only a few years later than they had anticipated. Not, however, later than Kurzweil anticipated.
The problem, he says, is in the nature of linear versus exponential projections. "An exponential projection is the same as a linear projection for a very short time." Draw both curves and in the early years, they look the same, but at a certain point, the exponential curve rapidly surpasses progress along a linear scale. To make matters worse, he said, an exponential curve can be subliminal to a linear curve for short periods, leading people to lose faith in their linear projections, as the more accurate exponential track slowly catches up. But it always catches up.
Kurzweil showed us a graph charting the development of practically every level of technology man has ever created, from fire to the world wide web. Even factoring in discrepancies between when various experts believe such things as the wheel were invented, the development of every single technological advance follows his exponential curve. It's not magic, just a vastly broad perspective.
"Six or seven years ago, most people didn't use search engines," he said, pointing to the latest in a series of technological advances. "The world is changing very dramatically, very quickly. The world will be very different in three to four years."
He then attempted to demonstrate the method of his particular brand of insanely intelligent madness. He told a story of talking to a business partner about devices that could be used by blind people to help them read signs and papers and menus in restaurants. The person wanted to know how soon digital camera technology could be used to make a device that would read the written word and convert it into the spoken word. Kurzweil had actually built such a device in the 1970s. It was roughly the size of a dishwasher - not very portable.
Kurzweil listed off the problems such a device would encounter, from trouble with lighting, to image stabilization, to pattern recognition, and surmised that it would be 2006 before the hardware was intelligent enough to deal with these problems. The year was 2002. But that wasn't the only obstacle; the software to interpret the images captured by such a device, and then accurately transmit the written text to spoken words hadn't yet been developed, and that would take about 4 years. The time had come, in other words, to force the convergence. That's what Kurzweil does. So they started working on making the device a reality. By 2006, he says, they had a prototype.