Semantic Web (Web 3.0) Will be new kid on block soon.
I've always thought of Web 1.0/2.0/3.0 in parallel to Tim Berners-Lee's notion of the read-write web, which is often used as an alternative way of describing Web 2.0.
First, we had Web 1.0 - the read-only web. Then came Web 2.0 - the read-write web - all of these services that make it easy for us to contribute content and interact with others. If you keep up the programming analogy, the next phase would be Web 3.0 - the Read-Write-Execute Web.
Most often, I hear people describing web 3.0 as Web 2.0+ - ie, more of what we have now, but faster and more ubiquitous. It's the same old stuff with lots of superlatives added. A Web 3.0 than plays on the idea of read-write-execute would be having a web that gives people the tools to craft their own tools, their own software, etc, rather than just uploading stuff to other people's software. Why go to YouTube when you can set up your own MyTube, if you will, with the same functionality but tailored to the specific purposes of your blog/community website/etc. People will be able to create their own complex online social media tools with a few clicks, but remain linked to other tools and other people through tagging, rss and all that good stuff.
1) Easier, cheaper, and more pervasive. Only a fraction of humanity has anything to do with Web 2.0. Others stay to the sidelines because they find the technology too confusing or expensive, or they don't see the relevance. Bring another billion or so people into Web 2.0, and Metcalfe's Law alone will make it a radically different phenomenon.
2) Always on, everywhere. We've heard (and written) this one for a while, but it's true: As Web 2.0 follows us every step of our lives--in some cases whether we want it to or not--and the dynamics change.
3) Controlling our data. In the next gen, we'll have developed all kinds of systems to wrap our personal data with various types of protection. Some will be shared widely, some narrowly, some not at all. And new systems of reputation and ranking should help us figure out which data sources to take seriously and which to shun.
4) In Web 3.0, the web will be programmable. Programmable for everybody! Here programmable does not mean mash-up, does not mean personalization through configuration. Hope to be able to show it by the end of the year.
