We’ve seen a decent amount of hype the last few weeks surrounding the launch of Wolfram Alpha, with premises of ‘search engine evolution’ and ‘Google killer’ tossed about. But I admit that I haven’t been paying very close attention – Google meets most of my search engine needs, the last big launch (Cuil) failed to make a dent, and Mr. Wolfram isn’t the most dynamic spokesperson. So, I was pretty surprised test driving the now-live Wolfram Alpha to learn it’s not really a typical search engine.
What I discovered is that W|A is essentially a a spartan (lifeless?) Mahalo or Wikipedia focused on collating “objective data” and performing calculations (he’s the dude behind Mathematica, after all) with the assistance of volunteer curators. There’s not much in the way of people (no ‘dave zatz’) or products (no ‘slingbox’), and WolframAlpha provides relatively few overt outbound links to the web at large – a few in the sidebar, but most are buried in a Sources pop-up. ‘George Washington’ links Wikipedia, ’12 Monkeys’ data is scraped from IMDB -why not just start these searches at the obvious sites anyway?
But where Wolfram Alpha really seems to excel is in providing cleanly formatted company profiles (TiVo v. Netflix above) and census-type data. In fact, all search or calculation results can saved as a PDF. In theory. Amongst the various glitches I ran into (some pics below), one was an inability to export data. Once these issues are worked out and in its current form, W|A seems best suited for student research. With perhaps less tampering or bias than you might find in a Wikipedia article.
But decide for yourself – give it a spin and let us know what you think.