I was made first aware of the WolframAlpha project by a couple of colleagues who were looking into artificial intelligence and were impressed with this new search engine that could understand questions and respond with natural responses. As of right now, all search engines have a tough time with translating natural language questions into data from analysis. A typical search can be tripped up with the choice of words used by an individual and may get the individual user incorrect results to their search query. This semantic difficulty can make a search for the correct information on search engine especially tricky for those who do not have the background in library and/or information sciences.
A real language (as opposed to a search language) method of finding information on a search engine would be convenient and would advance how the Internet could be used. This method would make WolframAlpha a “computational knowledge engine” for the Web. Because of its ability to go beyond the simple regurgitation of documents that contains parts of the answer. If Wolfram Alpha can compute the answer to a posed question in the time it takes a user to scan through the other open frames on their web browser, WolframAlpha could redefine how we interact with the Internet.
If we are to believe that the web is a conversation from one-to-few or one-to-some, then the design and organization prevent an interactive discussion between the two parties. Web artifacts are framed in a context in a similar nature to the traditional publishing of a book or a magazine. The difference is merely the addition of other forms of media (audio, video, animation). To create a conversation, one must have (surprising enough) two parties communicating with each other. Since WolframAlpha should have the ability to answer certain kinds of questions in a form that is close to natural communication, it could be the starting point to the “Web as Conversation.”
It is important to note that this is not the first attempt at such a service. Cuil was designed based on the same principle early and was considered broken as it could not handle the strain of a large group of individuals using the site at the same time. If WolframAlpha can support being Dugg or SlashDotted or even an onslaught by the Twit Army and can maintain the same level of service, it will be huge.
