I have said it before and I'll say it again, Microsoft's Pablo Castro is one of the few people putting stuff out there from Microsoft that really seem to "get it". He knows how people want their data (well, he knows how I want my data, and that's really all that counts, right?) and he seems to be on the same page as everyone else that I have spoken to as far as the whole REST thing. People want their data to be located at discrete, uniquely identifiable URIs. End of story.

In case you have been living under a rock, or you really don't care about Microsoft's "data in the cloud" strategies, Pablo Castro is the technical lead responsible for such gems as the ADO.NET Entity Framework and Astoria. Astoria is a project that wraps up an Entity Data Model in a WCF service with a uniform URI query format that allows for RESTful access to relational data via XML, RDF, or JSON.

The notion of the semantic web isn't really all that new, but it has been gaining a lot of momentum lately. The short story is that right now everyone is using the Web to publish and view human-readable content. What we look at on a daily basis is graphical, textual, and has animations, flash, whatever. The bottom line is that the content is human-readable. The semantic web pushes forward the notion that in addition to using the web for human-readable content, it should be used for data as well. The means by which the data on the semantic web is accessed is through raw HTTP, through a standard representational format like XML or RDF. It's a fantastic theory but I think it's too eutopian at the moment. I don't think that anytime in the near future the web is going to be flooded with this huge sprawling green field of RESTful services exposing POX/RDF data for the entire world to consume. Ths is where technology and business diverge. Technologically speaking, the eutopian vision of the truly semantic web is quite possible, and many people are working toward that goal right now. If you look at it from a business perspective, however, the outlook is a little darker. Bottom line is that people aren't going to embrace the semantic web until they can make money off of it.

Tools like Astoria are a fantastic tool by which we can expose data in a way that jives with the vision of the semantic web. The problem is that there are business concerns to exposing data on the web, not the least of which is of course -how do you charge people for that data? How do you make money off of exposing that data? The great thing about a semantic web and standardized data location and access methods is of course mashups. If anybody knows how to get at your data, and they know that your data is referenced in a way that is similar to the way in which Bob is exposing his data, etc - then everyone can consume everyone's data and the entire world enters a euphoric bliss of data consumption.

So what I see really happening is that corporations are going to take baby steps. Perhaps they will adopt "semantic web" style philosophies internally... hopefully they will even be using Astoria to expose relational models and helper methods on top of those relational models to allow applications within a corporation to consume data. In my ideal world, this is the way much of an organization's data is exposed internally. The clash between technological philosophy and real-world business practice occurs when you try and deal with how to authenticate access to your data, how you charge for your data, and how you license your data, etc.

The great thing about tools like Astoria for exposing the data and tools like Silverlight for rendering exposed data is that regardless of what the business people decide the future of the semantic web is going to be - you'll be ready. In that regard, as long as people like Pablo Castro are still allowed to make some decisions within Microsoft, we will still see a steady stream of good things coming - at least from the data team, anyway.

• Algorithms of the Intelligent Web