This is still my stuff.

I registered early and so got to see the Wolfram Alpha webinar presentation on Friday evening. The depth of the analysis and the breadth of the scope Alpha was apparently able to apply to any given field made me reel at the possibilities; the coherence with which the four key components of "curated data; computational algorithms; linguistic processing; automated presentation" (Wolfram's categorisation) had been married gave the impression of a genuinely capable successor to Google, not in the form of a better search engine, but of a more likely first port of call for genuine information. On seeing the demo myself, I realised not only why the questions Mr. W. addressed in the webinar painted a slightly disappointing picture of the actual information stored in Alpha - it relies on curated data, and obviously they've not got round to curating all of it yet, so it didn't know all the answers - but also why whether it knew about a particular topic was the wrong question. It will.