Stephen Wolfram on Wolfram Alpha Knowledge Search Engine

Wolfram Alpha will go public in May 2009. It has already been hyped as the first serious challenge to Google and traditional search engines. This may be true in the future but for now Wolfram has set his sights on making his computational knowledge engine do precisely what it claims to do - to turn the raw information on the web into real understandable knowledge.

Wolfram Alpha brings together two of Stephen's most famous products: Wolfram's Mathematica software and his work on cellular automata published in A New Kind of Science. Both are fundamentally about performing computations on formally defined structures - be they mathematical integrals or chaotic cellular systems. What would be really useful is to perform computations on knowledge structures as written in human language. After all, isn't this what computers were supposed to do?

This is where Wolfram Alpha is very different to our current batch of search engines. Rather than being just a sophisticated concordance of the web, this knowledge engine will turn plain text into human language structures, then convert these into computational language structures that can be queried, and finally turn the output back into a readable human form. Answering simple questions has never been so difficult!

Some techie websites have had access to this new Wolfram Alpha and are generally impressed, although keen to stress that this engine will have some limitations at first. The data it uses has been curated and will therefore expand only as fast as this process allows. Because of this, some think this is more likely to be a challenge to the likes of Wikipedia than to Google.

Now, this bit is impressive:"Another query with a very sophisticated result was "uncle's uncle's brother's son." Now if you type that into Google, the result will be a useless list of sites that don't even answer this specific question, but Alpha actually returns an interactive genealogic tree with additional information, including data about the 'blood relationship fraction,' for example (3.125% in this case)."

I, for one, look forward to tryin it out.