It's been hyped by some as a "Google-killer", and after a period of invitation-only testing today Wolfram Alpha is available to the unwashed masses. It's been described as an answer engine, as opposed to a search engine, and describes its long-term goal as making "all systematic knowledge immediately computable and accessible to everyone... to build on the achievements of science and other systematizations of knowledge to provide a single source that can be relied on by everyone for definitive answers to factual queries."

Not that I've been able to make much use of it so far. For example it doesn't recognise Scotland yet, other than as a historical kingdom that ceased to exist in 1707. (Insert your own joke here).

And following Sir Robert Worcester's presentation at the Centre for Journalism this morning, I've tried to interrogate it for current UK political polling statistics. No joy. I've also asked it to find web traffic statistics for UK national newspapers - again without success. I have a feeling early reviews - at least in this country - will share my sense of disappointment.

Still, as it develops and overcomes its severe US bias, we may find journalists gravitating towards Wolfram Alpha for some specific fact-checking.

One to keep an eye on.

Comments

I do think you expect a bit too much of a semi-beta phase project. Furthermore, I also think you've misunderstood the whole goal. I do agree, that the Scotland-failure is rather chilling, but comparable are rather ubiquitious terms like 'parliament' (which is a two-story building in Sofia?!) amongst others. It has time and space to develop, and advance its knowledge-base by either human input or computer algorithms (google-like 'maybe you meant this' when you spell something wrong is also lacking and surprisingly annoying).

But the point is to get answers to specific questions (like you mentioned at the end of the article - journalists using to check facts). Its database is currently limited to basic dictionary and ripouts of wikipedia (seem to be; even states so when checking 'used sources'). Expectations, of course, are high.

I've been following the progress of this site for a few months now, and I must say that at this point it is a long way from the Web 3.0 that it should be aiming for.

There are various problems. First and foremost, Wolfram's own catalogue of information seems to be far from complete. However, unlike Wikipedia (which many people cite as a better alternative), the engine is not dependant upon user contributions. User contributions which could potentially be wrong. Time must be allowed for Wolfram to compile information in its own computers, which is surely acceptable.

The engine seems to focus on quantitative information, rather than qualitative. All of the results that I have retrieved seem to be number-based, and/or strictly confined to categories. There is no room for supplementary information that is essential to an answer.

The engine needs the intelligence to understand that some information cannot be categorised or presented in graph form. Some answers can't be based around numbers.

However, with more compiling, I hope to see much more information in many more forms.

 

'Google-killer' Wolfram Alpha goes live. But doesn't have all the answers