A Computer That Can Answer Any Question – Wolfram Alpha

Imagine typing almost any question into a webpage or even voicing your question into a handheld device and instantly getting an accurate answer.  In a stunning development that has literally shaken the industry of information and search, computer scientist Stephen Wolfram has announced a computational knowledge engine that will attempt to do just that.  The announcement from Wolfram may signify the beginning of a game changing revolution in knowledge analysis and retrieval and is not to be taken lightly.

Many will claim that this is no big deal, pointing out that Google’s dead simple website for years has been answering our questions with ease, but this entirely misses the point.  As a former Google employee of 5 years I can definitely see a difference here.  Google finds webpages, videos, images, and several other documents that are related to our question, but Google does not attempt to answer our questions directly.  In contrast, Wolfram Alpha intends to directly answer our questions with a fact based answer in human language form.  Such a capability has been envisioned in Star Trek like scenarios and other sci-fi scenarios for decades.  Yet despite countless efforts from researchers and industry, success has remained elusive.

Wolfram Alpha’s product has not even launched yet (except to a handful of individuals that does not include us) but here we are calling it a game changer.  What is it that makes Wolfram Alpha so special?

Here is the answer:  Whether intentionally or not, Wolfram Alpha has suddenly brought the idea of an accurate, computerized question answering service into the consciousness of the information industry.  It does not matter whether Wolfram Alpha’s technology is the greatest thing since sliced bread or a complete flop.  What matters is that an idea that has been around for a long time has finally come of age.  Now the race is officially on to make it a reality.

Ever since Google’s very first days, founders Larry Page and Sergey Brin have claimed that search was a long way from achieving its true potential.  They stated that an engine that could take any question and instantly respond with the best possible answer was the ultimate goal.  Yet in a strange twist that highlights the narrowed vision that can often accompany success, Google seems to have lost sight of this ultimate goal.  Over the last ten years Google search has advanced by light years in terms of quality and breadth, yet by and large it is still a document search and retrieval platform.  Google is stuck in the search engine paradigm (not necessarily a bad place to be!) and has made little progress in advancing towards the question answering paradigm.

With Wolfram Alpha’s entry into the field everything has changed.  Within the next 5 years we will possibly see an avalanche of companies jumping into the field and progress could be extremely rapid.  Companies like Google will not stand still…they will either develop their own competing technology or acquire.

The timing is right for a working knowledge engine to finally arise.  Computation is fast and cheap, algorithms are more advanced, and several toolkits, api’s, and information archives exist to build upon.  Wolfram Alpha demonstrates this reality with its use of Mathematica, an “off the shelf” tool that is a cornerstone of its platform.

Now before we get carried away, lets make sure we don’t lose our perspective on the big picture.  Google’s search engine is not obsolete.  Google represents the document search and retrieval paradigm, and this paradigm is extremely powerful.  For many types of information tasks Google is exactly what we want.  Wolfram offers a new paradigm that does not kill Google, but rather one that complements it.  Google is especially good at helping us to perform research on broad questions where a specific answer is not desired, but rather where a specific document or series of documents is desired.  Question answering services, such as Wolfram Alpha, will be especially good at answering questions that have one true, generally agreed upon answer.

Question answering services offer strengths, as well as weaknesses, that will largely determine how and when they are useful.  At first the service may appear to be a novelty that we seek out for specialized situations, but over time the service will likely invade countless devices and platforms around us, allowing us to work faster and more effectively.  We can imagine an api that will allow any device or service to query the question answering service on demand.  Cars, appliances, social networks, and even search engines like Google will be able to access the question answering service, giving them the ability to provide factual answers to user inquiries as a subset of their overall function.

By far the most extensive and enlightening description of Wolfram Alpha’s technology, slated to become available to the public in May 2009, comes from a review written by Nova Spivack that was published on Techcrunch.  Spivack’s article is your next stop if you want to learn more.

Tags: question answering engine, wolfram alpha