I have a lot of respect for the folks at Gartner, but their recent report in which they support the term "Web 2.0" yet claim that the term "Web 3.0" is just a marketing ploy, is a bit misguided.
In fact, quite the opposite is true.
The term Web 2.0 is in fact just a marketing ploy. It has only come to have something resembling a definition over time. Because it is in fact so ill-defined, I've suggested in the past that we just use it to refer to a decade: the second decade of the Web (2000 - 2010). After all there is no actual technology that is called "Web 2.0" -- at best there are a whole slew of things which this term seems to label, and many of them are design patterns, not technologies. For example "tagging" is not a technology, it is a design pattern. A tag is a keyword, a string of text -- there is not really any new technology there. AJAX is also not a technology in its own right, but rather a combination of technologies and design patterns, most of which existed individually before the onset of what is called Web 2.0.
In contrast, the term Web 3.0 actually does refer to a set of new technologies, and changes they will usher in during the third decade of the Web (2010 - 2020). Chief among these is the Semantic Web. The Semantic Web is actually not one technology, but many. Some of them such as RDF and OWL have been under development for years, even during the Web 2.0 era, and others such as SPARQL and GRDDL are recent emerging standards. But that is just the beginning. As the Semantic Web develops there will be several new technology pieces added to the puzzle for reasoning, developing and sharing open rule definitions, handling issues around trust, agents, machine learning, ontology development and integration, semantic data storage, retrieval and search, and many other subjects.
Essentially, the Semantic Web enables the gradual transformation of the Web into a database. This is a profound structural change that will touch every layer of Web technology eventually. It will transform database technology, CMS, CRM, enterprise middleware, systems integration, development tools, search engines, groupware, supply-chain integration, and all the other topics that Gartner covers.
The Semantic Web will manifest in several ways. In many cases it will improve applications and services we already use. So for example, we will see semantic social networks, semantic search, semantic groupware, semantic CMS, semantic CRM, semantic email, and many other semantic versions of apps we use today. For a specific example, take social networking. We are seeing much talk about "opening up the social graph" so that social networks are more connected and portable. Ultimately to do this right, the social graph should be represented using Semantic Web standards, so that it truly is not only open but also easily extensible and mashable with other data.
Web 3.0 is not ONLY the Semantic Web however. Other emerging technologies may play a big role as well. Gartner seems to think Virtual Reality will be one of them. Perhaps, but to be fair, VR is actually a Web 1.0 phenomenon. It's been around for a long time, and it hasn't really changed that much. In fact the folks at the MIT Media Lab were working on things that are still far ahead of Second Life, even back in the early 1990's.
So what other technologies can we expect in Web 3.0 that are actually new? I expect that we will have a big rise in "cloud computing" such as open peer-to-peer grid storage and computing capabilities on the Web -- giving any application essentially as much storage and computational power as needed for free or a very low cost. In the mobile arena we will see higher bandwidth, more storage and more powerful processors in mobile devices, as well as powerful built-in speech recognition, GPS and motion sensors enabling new uses to emerge. I think we will also see an increase in the power of personalization tools and personal assistant tools that try to help users manage the complexity of their digital lives. In the search arena, we will see search engines get smarter -- among other things they will start to not only answer questions, but they will accept commands such as "find me a cheap flight to NYC" and they will learn and improve as they are used. We will also see big improvements in integration and data and account portability between different Web applications. We will also see a fundamental change in the database world as databases move away from the relational model and object model, towards the associative model of data (graph databases and triplestores).
In short, Web 3.0 is about hard-core new technologies and is going to have a much greater impact on enterprise IT managers and IT systems than Web 2.0. But ironically, it may not be until Web 4.0 (2020 - 2030) that Gartner comes to this conclusion!