The content of encyclopedias and other reference works such as dictionaries is essentially static. Of course, books have new editions and the content of Wikipedia and similar websites is very fluid. But when you look something up, what you see already exists. It isn’t generated dynamically.
WolframAlpha works differently. It takes your query and tries to use natural language processing to make sense of it. Then it gathers data from its own sources and from other sites. It then applies algorithms to the results, to calculate a dynamic response to your question. It describes itself as a ‘computational knowledge engine’.
In the next activity you will first use WolframAlpha to compare two kinds of dinosaur and then experiment with queries of your own.
Activity 8 Ask WolframAlpha
Go to thewebsite. You will see a search box and several clickable buttons. These take you to demonstrations of WolframAlpha’s capabilities (see Figure 10).
In the search box, enter
and press Enter. After a few seconds, a side-by-side comparison appears, with a table of data, images of skeletons, descriptions of the animals, and a graph showing how many times the related Wikipedia pages have been visited over time.
This is a relatively simple query though. More impressive is when Alpha uses natural language processing. Try these searches:
- how many words in Hamlet
- how many monsters in Loch Ness
- cells in a human body
- meaning of life
It is easily confused though. For example, it can’t really make sense of the question
why did the duck cross the road
Crossword and ScrabbleTM enthusiasts will find it useful. Try typing
Alpha is good at statistics and numbers generally, for example
- Olympic medals
- deepest lakes
- 2017 Roman numerals
If you explore Alpha further, you will find it is often surprisingly useful, especially in technical and scientific areas. But it is not an infallible oracle, and some of its interpretations are very wide of the mark.
The next section looks at cross-checking different sources of information and what can go wrong.