Sunday, June 03, 2007

Back to the Conference

SemTech finished over a week ago now, and I still haven't written all my notes down. Now I'm wishing they were more descriptive, as I just wrote down a few things to prompt me into remembering. Only I never thought I'd take so long to write it all out. If you ever need to set aside time for writing, then make sure you never have children.

Web 3.0

The big buzzword of the conference was Web 3.0. This seemed a little crazy to me. I think it's premature to create a label like this when Web 2.0 is still sorting itself out. In other words, it just sounds like marketing hype. On the other hand, all the CTOs, etc, seem to be lapping it up, so maybe they need this kind of hype.

Web 2.0 is about links and networking on the internet. "Mash-ups" are about someone manually making a link between resources, but it's also about emergent networks, particularly the kind where people are linking organically to each other. Web 2.0 is heavily dependent on people just doing their own thing, leading to something new that is an interesting (and sometimes unanticipated) gestalt. It's this pattern of emergent usefulness that makes Web 2.0 really interesting, but also explains why it is so hard for people to pin down. Look at O'Reilly's description and you can see that even the people who coined the term have trouble giving it an exact definition.

I may have completely missed the point, but Web 3.0 is basically Web 2.0 with semantic information built into it. While this sounds great in theory, I tend to agree with Google that most people aren't going to bother. In order to get "emergent" behavior or information from elements of the system interacting, you need for those elements to exist all over the place. If the process of putting these elements in is too difficult or time consuming, then it will never see critical mass. This seems to be the source of criticism from many quarters, and I tend to agree. Alternatively, you can try to automate the process with metadata extraction, but just as many people are critical of this too.

Not to say I don't believe that this can work. A simple example is seen in the "tagging" system in use on many sites. This is usually so easy that people are more than happy to take the few seconds necessary to click on a few items, making their lives, and the lives of those around them just a little easier. If we can provide more systems like this, then the idea might just work. But it's not a certainty.

While discussing my concerns, it was pointed out to me that there were a couple of companies at the conference who were referring to Web 4.0 and even Web 5.0. I heard a few of these justifications from some of the corporations involved, but we need to see how some of the earlier iterations pan out before we can determine the direction of the next steps.

In a quiet moment it also occurred to me that most people don't want to run .0 level software. It's too buggy. Maybe we should be aiming towards Web 3.1. :-)

1 comment:

Anonymous said...

Silicon Valley Entrepreneur & Strategy Consultant Sramana Mitra, after receiving reactions and feedback on her definition of Web 3.0, writes a follow-on synthesis explaining why the Semantic Web can only be implemented in a Contextual Domain. Thus, Web 3.0 according to her, is a Verticalized, Contextualized, Personalized Web.