Good day fellow Internet adventurers! I’m Ian. Marc has invited me to guest blog for you today. I am an Iowan transplanted to the east coast. My interests are broad, but include anything to do with technology, philosophy, or politics/economics. I also have a goatee—which qualifies me to think and speak on a range of subjects. Anyway, now on to the subject of this article.
Many of you have probably heard rumblings around the Internet about a new “search” site called Wolfram Alpha. Today, I will try to give you a short introduction and give you some ideas of what it is, what it isn’t, and what the future holds for the project. First, some background information. This project is the brainchild of Steven Wolfram who is the brain behind Mathematica—basically beautiful, and complex calculation and computation software. He also wrote the book “A New Kind of Science”, which attempts to lay out a new type of experimental protocol to make scientific inquiry more computational—and in the process wrote one of the best books on complexity in existence.
His latest project, Alpha, is to be his magnum opus. To understand Alpha, we first must understand why it exists. The problem of knowledge capture has come to the forefront of modern society. The problem isn’t that the knowledge isn’t there, it’s that there’s too much of it for any one person to feasibly tap into. Traditionally, when we want access to knowledge there were two ways of doing it. First, was to ask an expert who had previously sought out the knowledge you are looking for. The second was to acquire it for yourself with an exhaustive search through your library, or more recently, Google. This works out well if what you want to know has already been figured out, or someone has gone to the effort to put together a specialized calculator. This also means you may have to do some legwork to assemble all the individual pieces. I will use this opportunity to get into our first example. Say you wanted to know the value of a certain volume of gold. This is a rather straightforward sounding problem that takes a lot of legwork to solve. First you must convert volume to mass—not so hard if you have your periodic table handy. Then you must find the right units of measurement. In this case, it’s the Troy Ounce. After you have converted Kilograms to Troy Ounces, you must next look up the value of Gold per troy ounce as publicly traded on any given day (or at any given time). Then after you’ve done all that you have your calculated estimated value…most likely in dollars, so there’s another conversion if you want to know the value in some other currency. It doesn’t sound so hard on paper, but as good as I am with Google it still took several minutes to do the required calculations and conversions. Enter Wolfram Alpha. Let’s find out what 15 cubic inches of gold is worth in dollars.
Gee, that was simple. It took seconds. Notice 1) plain English search terms, 2) the option for conversions by the “current commodity price”, and 3) all the extra information provided. The rest is included below.
Also notice the toggle for temperature units. They also very nicely included down at the bottom, a link to source information. This is so you can verify the sources and make sure you are getting the right information. Furthermore, it allows you to generate proper citations if you wish to use the information in a research project.
It is important to note now, that Wolfram is still in its early stages. It relies on huge databases and powerful calculation engines (some from Mathematica, even) for its information so it cannot be relied on for perfect accuracy all the time—however this will improve as the information improves. It also has a limited range of subjects. For example, it isn’t so good with probabilities so far. The query below demonstrates one of the limitations.
As you can see it was having problems with the idea of probability, although it did try to direct me to other areas where it does have information related to my query.
The beauty of all this is that, this is only the beginning. Wolfram wants to make it so that eventually it will in essence be an artificial intelligence that you can ask big computational questions to and have it return a useful, concise answer. For now, though it is very limited to readily available data, and calculations.
Before, I mentioned it would not replace Google. Google is like a card catalog in a library. It indexes readymade sources of information and spits it out without any computation. Useful if you want to know the phone number for the nearest JC Penney’s, not so much if you want to know the P/E ratios for 4 different companies compared to each other. Let’s try another new search engine: Microsoft’s Bing.
This is the return we get, very similar to what Google would provide, anyway—not sure how it really advances the art of searching. Again, you could find each individually, but this is tedious work–Especially if you are trying to put together a large spreadsheet. Let’s see how Wolfram Alpha handles it.
Bingo! Sorry, Microsoft. All nicely formatted, straightforward, and free of extraneous information. The numbers are even aggregated in a table that could be copied straight to a spreadsheet.
As I don’t want to bore you with endless examples—which I could, I will end this article for now. For those of you who are starting to get the idea, you should explore the examples available and try some of your own queries. I may try to put together more demos later, especially if the talented and handsome blog-meister here at Teknosophy will have me back. In the mean time, however, I would like to invite questions, comments, and any stories of success or failure with Wolfram Alpha. I hope, however that what you walk away with in the end is the concept of how this changes the landscape for searching, not just for information, but also for knowledge.