Dan Gillmor has written a nice, accessible overview of the evolution of the web in his periodic column for the Financial Times. As he sketches it, vesion 1.0 was a “fairly static,” “read-only” affair – sites were relatively basic and we checked them for new content or downloads. Online retail and search engines sprang up, essentially to help us find things to read, while things like GeoCities made it possible for anyone to have their own site. With 2.0 it became a two-way street – a “read-write” web, with its poster child the blog. Now, we are learning how to weave all the pieces together and to recombine them in innovative ways – this is version 3.0.

The emerging web is one in which the machines talk as much to each other as humans talk to machines or other humans. As the net is the rough equivalent of a computer operating system, we’re learning how to program the web itself.

A big part of 3.0 are the “web services” that can be built with a site’s “applications programming interface,” or API. An API is essentially a window into a site’s code that programmers can use to build derivative applications. Google, Yahoo, Amazon, and Flickr all have APIs. Gillmor points to a wonderful site – ALL consuming – that uses the Amazon API to build communities around the media – books, music, film – that people are consuming. You simply post the latest entree in your media diet – anything that can be found on Amazon – and then add tags and comments. People inevitably find each other through what they are reading and discussions can ensue. This is an interesting step toward the real-time reading communities that will be possible when we have dynamic electronic books that can plug into the network.