Raw

John Markoff's NYT piece on Web 3.0 has generated a lot of response in the blogosphere, it's right on top of today's techmeme. Unfortunately (and predictably) its broad sweep across AI and the semantic Web [sic] has led to a fair bit of repetition of popular fallacies about the latter. Arguments are built on misconceptions of the Semantic Web, and from there the initiative is logically dismissed (yes, I am saying a lot of people don't get it - but I don't think it reflects on them).

An intelligent example comes from Ross Mayfield : 

I'd bet the future is less the Matrix than Soylent Green. Less semantic fuzz than social discovery.  Less artificial intelligence than human intelligence. Less automation and more augmentation.

I suspect he's right in emphasizing the social angle. But " semantic fuzz"? Whether " less artificial intelligence that human intelligence" makes sense depends on your definition of AI (and HI). The last sentence is a kicker. A significant promise of the Semantic Web idea is that by getting the machines to work with data on the Web more effectively, more automation will be possible - in the interests of augmentation.

Of course there are counter-posts from developers that are using these technologies ( teh community). Nova Spivack has a post on the misconceptions (including a couple of points with which I'd disagree, I'll get to those in a moment), with follow- ups on the topic of this post. Shelley has On Meaning and The Elephant Strikes Again. (I'm not sure I'd apply it to my discussions with Bill, but the "Elephant & Blind Men" phenomenon certainly does muddy the waters). But Dave Beckett wins the sanity prize with Semantic Web Hype (although there's a point in that I think might be debatable).

...we should show how you don’t need to boil the ocean of semantics such as applying a big pile of technologies or requiring something fragile like a single web-wide ontology - wrong, wrong wrong. Start from concrete data-centric approaches that build up to use layers of technology solutions to different problems as they emerge, only if needed and demonstrating usefulness at each stage.

I'll quibble a little: the only if needed part, because we don't actually need many of the technologies we currently use, they just make life easier ...demonstrating usefulness at each stage - hmm, I've got a load of wood in the chickenshed that took a fair bit of effort to obtain (reclaimed from wine barrels). I don't have an immediate specific application, but I'm pretty certain it'll be extremely useful sometime in the next year. I'm sure there are analogies with data representation. But whatever, I think Dave's approach should probably be the default.

Yaargh, gotta dash - but quickly on Nova's post," The Semantic Web is not just a single Web." - yes it is! (There's probably a touch of the elephant here). I'd also swap the " metadata" references to " data".

 

@en