WolframAlpha: much ado about nothing?
I finally found some time to try out the Internet’s newcomer: WolframAlpha.
I wanted to first try the system before expressing an opinion on it, so that I could based my judgment on real, concrete evidence instead of just on speculation and theory. I thought my expert eye would need at least three or four hours to be able to scrutinize the pros and cons of this service (yes, a “service”: although everyone tends to define it as search engine, Stephen Wolfram has never called it that). In reality, it took just a short time for me to realize that, at least for the moment, there is nothing revolutionary about this “computational knowledge engine”.
Basically, it is a system developed to obtain mathematical, scientific or statistical information (most of which is already available on the Web) in the simplest and fastest way possible, making everything accessible from one entryway. At the moment, the whole thing is quite second-rate, but this normal, as it is the very first version. I am neither passionate nor an expert of the content that the system makes most accessible, so I really don’t know which advantages WolframAlpha would be able to offer to who actually works on these subjects or just uses them for fun. The impression that I get is that the service is not of much relevance for the everyday user. I tried asking it a few questions in the more general dominions (such as sports and games for anagrams), and I noticed that the coverage and comprehension are seriously limited. Also, the module that works on the elaboration of natural language is quite primitive: considering that more than 200 people worked for over 2 years on this project, I think they could have done a lot better with this aspect (and I know how they could have :-).
Most likely, this service has fallen prey to the excessive expectations of those who gravitate around the world of search engines. The frenetic, almost messianic wait for the Web’s new Chosen One, the Google Killer, makes people lose sight of the fact that the creation of something which is actually innovative and useful for millions of people, is extremely complex and difficult. The fact that Google hasn’t presented substantial research innovations in years (besides some aesthetic touchups, longer abstracts and correlated keywords), goes to show that the road towards real improvement is definitely uphill.
If we get rid of all the hype which has built up during past two months, just a couple final considerations remain: first of all, it is a good thing that someone is trying to do something new (the efforts of Wolfram and his team should be praised); second, it may be best to wait at least another month before passing final judgments. The world is big enough to give room to successful services, even if they are aimed at specific sectors (in the end, the market is nothing else than a bunch of these sectors put together… ).
The only thing that I can say for certain is that the hunt for the new Google killer is still on, giving journalists (and me) something to write about for a long time :-).