You Cant Think About What You Dont Know

By Peter Coffee  |  Posted 2002-07-22 Print this article Print

Peter Coffee: In a world where too many companies (and their applications) work with too little data, Wolfram sets a good example with its latest version of Mathematica, which continues gathering new data to provide an evolving view of a problem.

It would be, at best, a gloomy satisfaction to be the person who wrote the report—released on September 11, 1998, if youre looking for painful ironies—that predicted a "catastrophic" intelligence failure if American agencies funding priorities werent substantially revised. Last weeks report from the House Subcommittee on Terrorism and Homeland Security found the resulting systemic problems, such as widespread shortage of foreign-language skills, to be much greater threats to our security than bogeymen such as off-the-shelf encryption.
I was struck by an associated story on National Public Radio, which noted that headquarters counter-terrorism staff had been expanded since that time—but that this larger analysis group had less raw material to analyze, thanks to a concurrent reduction in field intelligence collection assets during that same period. More analysis of less actual data? This may be the defining syndrome of the era of cheap computation, as people are tempted to do more of what costs less with every step down the Moores-Law curve—while cutting back on the old-fashioned efforts that seem ever more "inefficient."
Its nice to see an exception to this trend in the latest release of Mathematica, version 4.2, from Wolfram Research, in Champaign, Ill. A tool like Mathematica could encourage the introspective, "Im sure I can figure this out myself" approach that is embodied in company founder Stephen Wolframs controversial book, "A New Kind of Science," now topping every true geeks summer reading list. But even while Mathematicas continuing computational refinements enable ever-deeper exploration of an isolated idea, the XML facilities in the new release—combined with the XML APIs now being exposed by many Web sites—also make it a tool for going out into the world and collecting new raw facts. For example, during a conversation last week, Wolframs Director of R&D, Roger Germundsson, showed me that he could write an "AmazonSearch" function in Mathematica that would drop a formatted table of authors and titles meeting certain criteria into the middle of a Mathematica "notebook" document, querying the XML-based interface that Amazon makes available for use by its retail partners. Its also important that Mathematica 4.2 emphasizes enhanced tools for publishing its results, as well as for collecting initial data. For a counterexample, look at the way that the House subcommittee formatted the online report that I hyperlinked above: Its a Word file, but all it contains is images of printed pages. Terrific: You need Word to open it, but it cant be indexed or searched or readily excerpted. How many ways can we do this wrong in a single, simple document? Applications need continuing access to new data if theyre going to do anything more than confirm our initial hypotheses. Analysis needs to be shared, in ways that both expose our assumptions and make our results available as catalysts for others thinking. And enterprise applications, such as those for relationship management, need to avoid making the mistake of thinking that more analysis can ever make up for having less, or lower quality, information from the real world. Tell me how you keep your applications well informed.
Peter Coffee is Director of Platform Research at, where he serves as a liaison with the developer community to define the opportunity and clarify developers' technical requirements on the company's evolving Apex Platform. Peter previously spent 18 years with eWEEK (formerly PC Week), the national news magazine of enterprise technology practice, where he reviewed software development tools and methods and wrote regular columns on emerging technologies and professional community issues.Before he began writing full-time in 1989, Peter spent eleven years in technical and management positions at Exxon and The Aerospace Corporation, including management of the latter company's first desktop computing planning team and applied research in applications of artificial intelligence techniques. He holds an engineering degree from MIT and an MBA from Pepperdine University, he has held teaching appointments in computer science, business analytics and information systems management at Pepperdine, UCLA, and Chapman College.

Submit a Comment

Loading Comments...
Manage your Newsletters: Login   Register My Newsletters

Rocket Fuel