Science Reporting Goes Fzzzt

Rather than provide my own break down of John Markoff's (real name!) article in the New York Times, I'm happy to point to Greg's post which captures the right level of frustration. I'll add, though, that the article naively throws together a number of ideas (mixing interpretation of data, with reasoning on or about that data), some timelines ('the effort is in its infancy' - AI is not in its infancy, we have been engaged in this 'effort' for decades) and general confusion (social tagging like Flickr has nothing to do with AI) to create what seems like a make-news story aimed to leap-frog discussion around the Web 2.0 festivities in California last week. Finally, things like this:

... Web-based e-mail programs recognize dates and locations. Such programs, the researchers say, may signal the impending birth of Web 3.0.

Indicate that the reporter isn't really in touch with NLP, AI or other research efforts.