Wolfram Alpha is a site developed by Stephen Wolfram. Leo Laporte conducted an interview with Mr. Wolfram on Saturday May, 16 2009 amidst the launch of Wolfram Alpha. The interview was originally included in this episode of the podcast TWIT. This segment of the podcast was extracted and the interview can be found here:
People have confused this site as a Google competitor. It is not. While Google’s mission is “to organize the world’s information and make it universally accessible and useful”, Wolfram indicated that their goal is “making all the world’s knowledge computable.” This endeavor required a long term effort to develop a language that allows representation of all kinds of knowledge including all the world’s algorithms.
Wolfram explains that even though the world appears to humans as a rich and complex environment, in reality, it is built on simple programs and frameworks. It was this realization from his scientific background along with the comprehensive language that allowed Wolfram Alpha to be built.
The concept of Wolfram Alpha is based on a computational approach to science with the idea of exploring the universe of possible programs. Current programs are built to achieve a specific purpose. Wolfram’s concept can be illustrated by the concept of taking a simple program and enumerating all possible programs of this kind. He indicates that most of the programs would be uninteresting and useless, but that it is possible that some will behave in interesting ways. If the simple program produces patterns based on previous output, then if you enumerate all the possibilities, you would end up with some interesting patterns.
Wolfram indicates that this is the way that nature behaves. He indicates that nature might be doing a random sampling of the programs available and this is what creates interesting patterns. He contrasts this a bit with Natural Selection by saying that Natural Selection would indicate that a current form is a history of all influences on that form. Wolfram continues that he theorizes that nature randomly selected simple programs with complex behavior which generated the form. Specifically, he uses mollusk shell patterns. There were some that are simple such as dots or stripes produced by simple programs and some that are complex. He asks the question did the pattern “get sort of molded by some elaborate series of events in the history of Natural Selection or is it that out of the space of possible programs that is just so happens that there are a few that sort of spontaneously produced this kind of complexity?” He continues that he believes that it is more of the latter.
Wolfram believes that it is possible to enumerate all programs for various disciplines. He uses examples of axioms or mathematics or laws of physics. He continues that you might be able to “find technology” by enumerating possible programs and algorithms. One of the tasks that Wolfram Alpha performs is algorithm discovery.
He indicates that people think of Biology as a very accidental science which is the result of the history of a series of accidents and people think of Mathematics oppositely and are built on higher principles. He believes this situation could actually be inversed. He believes the mathematic and logical axiom systems are purely historical and there are others out there that we could have studied, but previous cultures settled on these approaches for study.
He indicates that Wolfram Alpha is an application of this paradigm of this science of making a lot from small principals using of the historical facts, models and methods of science, financial analysis, physics, etc. He believes that the potential great use of Wolfram Alpha is the development of new models to answer new questions to which there have been no previous answers and is it possible to mine “on the fly” from the computational universe and build something that has never existed before?
Wolfram contrasts this approach with Artificial Intelligence which traditionally has attempted to emulate human reasoning and automate it by explaining that prior to Newton, people would debate the philosophy of an idea. Then Newton showed that debate was unnecessary and one could simple “blast” their way through with mathematics to arrive at the answer. Wolfram states that his approach is an extension of this by enumerating the options and just “blasting” through to the results.
He continues to explain the way Wolfram Alpha works at a mechanical level. Basically data comes in and is parsed to understand it. Once understood, the system determines what information and data has relevance to the input. Finally, these results are considered in terms of what is the best way to represent the information and then presented.
What application does this have in the Medical Community? The goal of a Cybernetic approach to medicine is to insure the desired outcome from a quality standpoint. Clearly, we are on the cusp of Computing Science’s ability to assist in this endeavor. While EMR/EHR systems capture the data and provide some rudimentary assistance, we must demand further capabilities such as automated second opinions, built in decision support and allowances for technology infusions of the type described above. It will only be these types of additional tools which provide physicians security to execute with assured outcomes as patient loads increase and complexity deepens.