Wolfram Alpha launched last weekend to much fanfare. In fact, this scientific search engine received the most hype I can recall for a tech launch since Dean Kamen's Project Ginger (aka the Segway). Of course, today, we're all navigating the sidewalks on our Segways and cities have been "architected around the modern transportation system" (as once predicted by Steve Jobs), so in a few months most of us will barely remember what Google once did.
Sarcasm aside, Wolfram Alpha is an interesting application. Characterized as a "computational knowledge engine", Wolfram Alpha aims to compute answers based on facts stored within its knowledge base. To some extent, it is more Wikipedia than Google.
Its greatest strength is in performing calculations. While Google can do this as well (provided that the formulas are structured correctly), Wolfram Alpha can perform more complex calculations and seems to do a better job at recognizing various constructs of a formula. That's not surprising since Wolfram Alpha is built on top of the Mathematica software created by its founder.
But Wolfram Alpha is largely constrained by its knowledge base. It's as much a content company as a technology company, and as such, its ability to answer questions is based upon the depth of information in its knowledge base.
Beyond mathematical formulas, the Wolfram Alpha knowledge base seems deep in scientific knowledge, not surprising based upon the pedigree of its founder. You can compare the mass of two planets, compute the airflow around an airfoil or compute someone's body mass index.
For most users, however, Wolfram Alpha seems likely to disappoint. One clue to this is the sample searches they present on their home page - enter your date of birth or the city in which you live and Wolfram Alpha returns a list of people born on the same date or basic facts about your city, derived from the CIA World Factbook. In fact, for non-computational queries, Wolfram Alpha seems to me like an online almanac. The last time that I looked at an almanac was as a child, so I'm not quite sure why I'd want one now, but the next time that I need to know what the weather was on the day I was born, I guess I can turn to Wolfram Alpha.
My challenge in using Wolfram Alpha is that there are already individual sites that provide me with most of the answers it can offer. Today's weather in London? I can find that on Weather Underground. The distance between Boston and Cape Code? Google Maps or Mapquest will suffice. Wolfram Alpha consolidates a lot of that information (and might make for a useful iPhone app) but it's hardly the game-changer that it has been hyped to be. So, when I need to compute binomial coefficients, I may jump on my Segway and fire up Wolfram Alpha, but for most of my day-to-day search needs, I'll stick with Google.
