Do you remember the cryptic puzzles that would ask you to multiply the Wonders of the World by the Heinz varieties, then subtract the horsemen of the Apocalypse?*
In a nutshell, that is Wolfram Alpha. This 'computational knowledge engine' turns plain English queries into hard data, computing numerical answers from common sense questions.
Of course, it's not quite that simple. Wolfram Alpha doesn't know how many kinds of ketchup Heinz makes, nor whether to include the Great Wall of China in a list of world wonders.
What Wolfram Alpha does know is over 10 trillion pieces of factual scientific, mathematical, socio-economic, biographical, cultural and linguistic data, and - roughly at least - how they relate to one another. So if you type in "depth Loch Ness / height Nelson's Column", you'll discover that you could sink 2.5 columns before hitting Nessy.
Wolfram Alpha's knowledge is impressive. You can drill down from national mortality rates to see where you fall in a distribution curve of height or weight, then further to calculate the calories you'll burn from your daily jog, then to the nutritional content of your breakfast and finally even examine chosen DNA sequences, bio-chemical reactions and particle physics.
In the other direction, you can pull up technical details on bridges and landmarks, then see routes between cities on a globe, including population and timezone data.
Expanding your focus, graphs of weather scale up to the paths of future solar eclipses, the actual distance today between the Earth and the other planets, interstellar astronomical data and ultimately, particle physics again.
With a bit of practice, you'll soon be able to calculate how much energy it takes to boil the world's oceans, compare road traffic accidents statistics worldwide and plot your choice of earthquakes and hurricanes - complete with official and academic sources.
And that's before you even start tackling pure mathematics. Wolfram Alpha makes tackling advanced formulas so easy - even walking you through step-by-step solutions - that I tremble for the intellectual future of university students.
We guarantee Daily Mail-style outrage over the way it breezes through the toughest polynomial, integral and differential maths homework.
But a Google-killer or a Wikipedia-slayer, Wolfram Alpha is not. For anyone used to the millions of results Google spits back, Wolfram Alpha's reticence to commit to an answer could be frustrating.
The natural language engine's comprehension levels are variable at best, with the same words being understood in different ways when you string queries together. Unless you're very precise - and rather scientific - in composing questions, you'll get very familiar with the response "WolframAlpha isn't sure what to do with your input."
I found that at first about three quarters of my queries were rejected - and even after a lot of practice, about half still failed.
Similarly, its responses to encyclopaedic queries such as 'cheese' or 'World War 2' are very brief, very dry and utterly unaffected by topical or popular issues. 'Jade Goody' pulls up her birth and death dates above a rather insensitive 'timeline', while the lack of any response for 'swine flu' illustrates just how tall Dr Wolfram's ivory tower is.
Or check out the result for 'h1n1', which brings up a beautiful 3D contour plot of the definite integral over a disc of radius R, namely integral_(h1^2+n1^2
But Wolfram Alpha is not pretending to be a Google or a Wikipedia. Its content is peer-reviewed rather user generated, and there is nary a whiff of commerce to be sensed. As a tool for students, scientists, researchers, teachers and journalists, Wolfram Alpha is the first flickering of a mature, reasoned approach to knowledge that deserves every encouragement.
There is also room for it in the mass market, as a trustworthy conversion engine, a reliable calculator and, hopefully, an increasingly sophisticated oracle for the world's exponentially growing scientific knowledge base.
Wolfram Alpha can be found at www.wolframalpha.com, and should be launching to the public on Monday May 18.