As someone involved with the Semantic Web (I'm a developer on the Semantic MediaWiki project), I was naturally quite curious about this book. There's been so much vague, buzzword-y and contradictory hype about the Semantic Web that the time is ripe for a book that cuts through the BS in clear language, and still manages to make a compelling case for the Semantic Web. David Siegel's "Pull" is not that book - it ignores the ambiguity that currently exists, and adds some more of its own. This is a book that often sacrifices clarity for hype.
Despite the subtitle, the basic idea of "Pull" is not actually about the Semantic Web at all: it's that, in the future, the services and products we use will know everything they need to about us - so that, for example, when we enter a hospital, the systems there will already know our complete medical history. In the book's parlance, each device or process "pulls" the necessary information to it, rather than requiring us to "push" the data - and the source of the information will be some sort of personal online "data locker" that each of us owns. It's hardly a new idea - variations of it a staple of speculative magazine articles for maybe 20 years; and I'm even aware of some failed '90s startups that tried to do a subset of it, like online "agents" that make purchases for you. Which is not to say that all of this stuff won't happen, of course; but it's an indication that there's no guarantee that any of it will come any time soon.
The new idea in the book is that the Semantic Web will be the thing that brings us to that point. But the way Siegel defines "Semantic Web", that's basically a tautology. He uses the term to refer to any set of data that is contained in a standard, non-ambiguous format, and is available online. Well, if you have distributed smart devices and services, they're going to have to send data back and forth with other systems - and if so, the data would have to be in a standard, non-ambiguous format, and sent over a network. It's really just a restatement of one of the challenges involved (though hardly the only one). And the concept of more-universal formats for communication, too, is a fairly old one - the book's description of a sort of global language of data closely resembles a lot of the hype around XML that emerged in around 1997. In fact, some of the projects Siegel mentions simply *are* XML, like XBRL, the business-based XML format that makes financial reporting easier to process. It looks like a case of existing technology being rebranded to fit the new buzzwords.
To add further to the confusion, some parts of the book have nothing to do with either "pull" or the "Semantic Web", but seem intended just to champion Siegel's pet causes, or maybe just to pad out the book - like his praise for replacing the income tax with the FairTax national sales tax, and his discussion of "robotic ants" that, in the future, will crawl around your house to check the wiring. The connection seems to be that, like smart devices, these innovations will make your life easier - but if so, maybe he should have just called the book "Ease".
For a book as nebulous as this one, it still manages to get a lot of technical details wrong. To give two examples: it confuses "data" with "metadata" - a pet peeve of mine; contrary to the book's description, a business card doesn't hold metadata, just regular data. And the book mentions my project, Semantic MediaWiki, as well as a somewhat related project, Freebase - but only for long enough to refer to them both - incorrectly, in both cases - as "Wikipedia-based efforts".
"Pull" quotes from Bill Gates' 1995 book "The Road Ahead", and one of the review blurbs compares "Pull" to that book - which is unintentionally appropriate here, because "The Road Ahead" is a cautionary lesson in the dangers of techno-futurism. Gates' book covered some of the same ground as "Pull" - smart houses and all that - but gave only brief mention to the World Wide Web, at least in its first printing. It's only a year after "Pull" was published, and already one can see it headed for a similar fate. Siegel advises Apple to get out of the computer-hardware business and into the data business: this was four months before the launch of the iPad. He also writes, "I expect Google Squared and WolframAlpha to be quite popular by the time you read this." Both currently languish in (semi-)obscurity, after some initial buzz. And he lavishes praise on other projects that have failed to get much excitement elsewhere, like NEPOMUK, LarKC and UMBEL. If you're reading this review more than a year or two after the time of this writing, it could be that one or more of those names have become a bigger deal - I doubt it, but unlike Siegel, I don't claim to be well-acquainted with the future.