Web 3.0

My Facebook news feed was cluttered with mockery this week after Walt Mossberg and others at the All Things D conference started throwing the term "Web 3.0" around. Walt defines it as "the thin client, running clean, simple software, against cloud-based data and services," but that doesn't seem to me such an amazing new development, nor am I sure that era is ever really going to arrive. In a world of mashups and digital art, nobody's going to be giving up Photoshop or ProTools anytime soon. And your wonderful iPod won't do much without iTunes to load it.

In any case, "Web 3.0" is just impossible to say with a straight face. Even "Web 2.0" is a pretty irritating buzzword, but it had the advantage of describing a very real and very fundamental change in how we used web content. The first generation of the web was about pages and web sites. Back in 2005 when I started working with the Ottaway local newspapers, I raised some eyebrows by saying their home pages were not that important, and by saying my favorite must-read web site was a blog aggregator. Once we got analytical software installed and running and editors could see how many article views were coming through search engine result pages and RSS feeds, they began to understand what I meant.

So "Web 2.0" was about a shift from sites and pages to widgets and components, where the "site" was broken up, mashed up, remixed, into contexts designed by the reader rather than the site creator. The true "Web 3.0" will be the next step in that evolution, where the components are broken down and the feeds themselves are ripped apart and remixed. It's not here yet, but it's on the way, and it's called the "semantic web." Every time you search for tag clouds on del.icio.us, or search Twitter for live data on particular topics, you're using the beginnings of the semantic web.

What does this mean for site creators and content management? Think small, not big. Write 20 blog posts instead of one big article, and make sure each one is loaded with metadata -- geocoding, subject coding, entity identification. Then think about what you could do with that data. Another way to think about this is that we're letting computers handle the dumb things (assembling pages, even creating feeds) while we humans do the things computers can't: tell others what this is and why it matters.