A 2003 post written looking at the shortcomings of the Semantic Web:
Excellent read, with thoughtfulness and insight into why the classic Semantic Web won't work for most. My talk at DERI covered a aligned topic of slow adoption , outlining the need for the Semantic Web more usable for web developers.
5 years later...
Much of his criticisms are against AI, description logic and the way data is interpreted. I agree with his outlook on those topics. I also agree with the need for seemly conflicting world perspectives, the difficulty in getting people to agree on standards, and the tremendous value of generalities in models, (which take a long time to build in "bottom-up" style AI efforts).
But it misses the true tact of the current semantic web - which is focused on three key areas:
Shirky's argument is mostly academic. And the thing is, academically speaking - he is right. However, as a recent Nova Spivak post points out - there's plenty going on at places like DERI. And the concerns Shirky raises are fast disappearing as collective intelligence addresses the weaknesses of the original proposals
About Clay Shirky and his recent book on the the Colbert Report.
Perhaps he has more recent thoughts which I haven't read about. There was an interesting almost immediate response.