I would agree that data representation is absolutely the key issue. We should remember however that data representation is not just about the data itself but how given a particular set of facts one can derive further facts by inference from those facts and furthermore how you can guarantee the consistency and integrity of this data (really a prerequisite for derivation to begin with - from inconsistent premises you can prove anything).
The future lies with methods of data representation that allow this. The semantic web crew are totally on the wrong track with their enthusiasm for XML - XML is a throwback to archaic hierarchical representation methods - already proved unworkable in the 1970s.
The relational model is definitely the right direction but we are badly let down by current implementations.
Time to say loudly: if a product describes itself as relational but still uses SQL it doesn't do what it says on the can!
Nick Carr is welcome to stake a claim for the trademarks on Web 3.0, but I can cite prior art. Judging by the reaction of the blogosphere over the past 24 hours, there's not much value in the term anyway. People are feeling deflated enough about Web 2.0 and have no appetite for yet another spin on the versioning meme.
So we can be fairly sure that the successor to Web 2.0 won't be called Web 3.0.A successor will rise out of the ruins of Web 2.0 There will be a successor, though, because Web 2.0 has to be remade before it can deliver on its promises. Perhaps that's at the root of the distaste for talking about Web 3.0. There's a growing realization that Web 2.0 is flawed. Web 3.0 — or whatever we end up calling it — will rise out of the ruins of Web 2.0.
Over the past couple of weeks, I've been thinking about the ways in which Web 2.0 will have to be recast. This is not an exhaustive list, more of an introductory post to raise some issues that I'll return to and expand on in future postings.
Views not data
I think Tim O'Reilly's assertion that "data is the 'Intel Inside' of Web 2.0" is misleading. It's a bit like saying, "Bits were the Intel Inside of the PC." They weren't actually, it was how they were processed that made Intel the platform of choice. Similarly, it's not the data itself, it's what you do with it that counts in the Web 2.0 era (which is why John Markoff's New York Times article is on the right track when he associates the semantic web with Web 3.0, even though it's only part of the story). Value comes from the views that you create to filter, join and represent data — whether it's your data or someone else's (more often the latter).
Production not friction
I think Bill Gates is absolutely right to criticize the mindless traffic-chasing that characterizes a lot of Web 2.0 ventures. People who think that building communities — or social networks as they like to call them these days — is alone the passport to wealth are missing the point. You can generate a lot of friction and heat by gathering lots of people together, but you only generate value if something gets produced that someone else wants to buy. Sometimes the serendipity of undirected friction will be enough to produce a result, but most of the time that productive effort needs to be directed in some way.
Guarantees not assumptions
Of course if you're going to sell something then your customers expect some quality control. So far, Web 2.0 has largely happened on faith. Mashups don't come with service level guarantees. Making sure that they do opens up some very complex questions about how to distribute quality of service responsibilities across multiple tiered shared services.
Promotion not advertising
It seems to me that advertising is increasingly irrelevant in a world where it's possible to embed service options in a process. Why advertise when you can present your service directly to your prospect at the point of need? Contextual advertising is just a starting point. The ultimate destination is contextual embedding of the services themselves.
