Thursday 28 May 2009

Wolfram Alpha -brilliant, but not quite yet


Having listened to the hype around the 'Google Killer' Wolfram Alpha, I was downright sceptical. Prior to its launch I even posted a discussion on the KIN Members' forum suggesting that its founder, the British mathematician Stephen Wolfram , was either a genius or fraudster. I now acknowledge that Wolfram is the former. What he has achieved is remarkable; the results of scientific based queries such as comparing species or resolving formulae are impressive.

My hope was that this would be leading the way for a semantic engine to make sense of the deep web and linked data, as Tim Berners-Lee has predicted. This may still be the case; the issue is not what it does, but what data it makes use of. It is therefore deeply unfair to position Wolfram Alpha as a 'Google Killer'. It is not using the web as it's information source. The data, unlike Google's raw material is well researched and verified and therefore bounded. I have no doubt that Wolfram is looking at both how to apply Alpha to the web, and also how to monetise it. When he figures that out, Google will be worried. When Wolfram Delta 'Corporate Edition' is available, connecting knowledge across the organisation becomes a reality.

No comments: