One of the foundations of the Singularity is the concept that technology is self-accelerating and that at some point it will become self-improving. I have not seen the self-improving part so much, but over the last few years I have seen the self-accelerating part.
When I started in my current field, GIS (Geographic Information Systems), in 2002-2004, digital geography was still in its nascent form and it was being utilized in a powerful, but not world-historical transformative scale. Its strange, I come into it at a time that was still relatively innocent, when you could put together a digital map and feel relatively impressed with yourself. That early experience has given me just the right amount of perspective for what has followed. I think this podcast (scroll down to the 2004 area and look for David Rumsey) by David Rumsey captures the pre-2005 mood perfectly. GIS was an exciting field and there had been many efforts by many devoted people to spread the information to the general public.
Then, Google Earth arrived and everything changed. Virtually overnight everybody's connection to geography changed and the expectations went through the roof. You can still impress people with a nice, clean digital map, but it is an order of magnitude more difficult to find such a person.
However, the scary thing is that the changes are just beginning and the advance of GIS technology is far out-stripping the ability of most people to capitalize on it (this applies to me and the people I work for and with). The revolution is only starting. There is a convergence of technologies occurring that will totally transform our relation to our world within the next decade. Many people have a limited exposure to these technologies through Google Earth and Microsoft Virtual Earth, but these only offer a small peek of what's to come.
Today, I listened to an intriguing Long Now podcast by Paul Saffo, a forecaster associated with Stanford University. In the podcast, Saffo mentions that the 80's were the decade of cheap processing power (the advent of the personal computer), the 90's were the decade of the cheap laser (the advent of fiber optics and the networked world), and this decade will be the decade of the cheap sensor. He right on because I've seen it in action in GIS, real-world (not quite real-time) sensing has gotten much more affordable. That's why Google and Microsoft have been able to invest in it on a Nationwide scale that the government could not even have imagined.
Combined with computer processing power and networking power, the ability to create models from this newly available sensing data has grown exponentially. It will not be long before we will be able to view a functional mirror world model of the real world that is a near perfect representation of the real world.
There are five key technologies involved in this transformation, these include LiDAR (laser-based sensing of the Earth's surface), hand held GPS, street-view GPSed photography (now viewable on Google Maps), cheap digital color orthophotography (straight down view), and oblique photography (the "bird's eye view" now available on Virtual Earth). While all these technologies have a certain "wow" factor on their own, the meat of them is the advancing, but still infantile, ability to pull these technologies together and create models. One demonstration of a model I recently saw was breath-taking.
As impressive as these technologies are, the amazing part to me is the utter lack of applicability these things have to most people in comparison to the richness present in the data. The data is extremely valuable, especially in comparison to price, and there is money to be made by the people who really figure out how to make it applicable to the average person.