One to Watch?

An item in yesterday’s Bits (“Better Search Doesn’t Mean Beating Google” NY Times – Technology) talked about a search engine that promises to do things differently. Wolfram|Alpha, the creation of mathematician Stephen Wolfram due out in May of this year, has been much ballyhooed as using “a mixture of many clever algorithms and heuristics, lots of linguistic discovery and linguistic curation” ((Wolfram Blog)) to provide answers to your questions, rather than lists of websites that relate in some fashion to your search terms.

Apparently, Wolfram has been making strong claims for the novelty and ingenuity of the process that will underlie a search in his engine, claims that some consider, well, unseemingly boastful and overblown. Whatever the case, it will be interesting to try it out when it launches; accurate answers would be a boon, of course, should it succeed, and if it fails, Schadenfreude will provide some small measure of pleasure ((See, e.g., the disastrous launch of Cuil.)).

Of course, it’s unclear whether Wolfram|Alpha aims at delivering up answers across a broad spectrum of information, or whether it is more likely to hew to the scientific and mathematical fields from which Wolfram comes.

Wolfram|Alpha is not alone in providing answers rather than leads to them. One (critical) article pointed readers to START, a “natural language question answering system” out of MIT. This is a handy addition to your long list of research resources, perhaps, although when I asked it who Canada’s first prime minister was, it first apologized and said it didn’t know, but on a second attempt it offered me general material about Canada’s executive branch… Queen Elizabeth, Stephen Harper, and all that jazz.


  1. What a huge information literacy challenge! Especially if they point this thing at “artsy”, non-mathematical/non-scientific information. At least with a hitlist of websites, I can judge the most qualified sources and use them. If this engine just returns “the answer”, I’m abdicating the weighing of the source to someone else.

  2. I’ll make sure to co-locate my Wolfram and Cuil bookmarks…