By Thomas Scott
Supercomputer learns then un-learns Urban Dictionary
IBM’s Watson made waves after its debut in February 2011 on the iconic game show Jeopardy.
But recently the supercomputer has taken to memorizing the entirety of the notorious slang database Urban Dictionary.
Watson is a super-computer designed by IBM to master natural language processing, a field that involves using artificial intellgience to facilitate communication between computers and people.
According to The Atlantic, researchers discerned that use of “informal language might be a great way for Watson to understand the way real people communicate.” But once Watson had successfully copied Urban Dictionary’s contents, it could not differentiate between appropriate language and profanity.
IBM’s website claims that Watson can “analyze the meaning and context of human language.”
However, researchers were forced to remove any trace of the site from Watson’s memory banks and devise a filter to censor any foul language.
The Artificial Intelligence Platform is being implemented in more practical applications, such as finance and healthcare, which deal with immense quantities of data.
IBM described progress in a March press release, saying, “Extracting medical facts … buried in large volumes of data” from a variety of sources like “electronic medical records, family medical history and the latest clinical research.”
But electronic records have not been as widely utilized as expected, which could perhaps impede Watson’s research goals.
A 2005 report by the Research and Development Corporation (RAND), a non-profit think tank, predicted that digitization of medical records was an inevitability. However, recent reports have elucidated just how much the adoption of electronic medical records have faltered.
According to Health Affairs, about 30 percent of U.S. hospitals have not made the prescribed transition.
According to an IBM press release, researchers are also seeking to apply new technology garnered from the development of Watson to “improve diagnosis and treatment in areas such as chronic disease and oncology.”
Watson has also peaked the interest of Citigroup, one of the largest financial firms in the United States. The multinational bank’s current level of involvement with Watson is exploratory.
Citigroup is interested in how Watson can aid them in evaluating “customer needs and process[ing] vast amounts of up-to-the-minute financial … data,” according to a Citigroup press release last March.
Citigroup is also interested in how the implementation of natural language processing can “help advance customer interactions.”
In 1954, Citigroup petitioned IBM for an electronic brain that required only nine and a half minutes to perform a cost-benefit analysis instead of 1,000 man-hours.
At the University of Rochester last May, an academic case competition was held. Entries were submitted by students of various disciplines within economics. The competitors were charged with devising ways to deal with “complex challenges in the transportation, energy, retail and public sector industries,” according to a University of Rochester press release.
The first prize was awarded to an entry titled “Managing Data in the Eye of a Storm.” The primary aim, detailed in a May IBM press release, was for Watson to combine “weather-related data and the latest census numbers to help organizations better prepare for a crisis administration” due to “capricious weather activity.”
Second place, also detailed in an IBM press release, went to a solution which sought to aid energy companies by “improve[ing] the understanding of environmental impacts … to reduce accidents while avoiding the over exploration of natural resources.”
For more information on IBM’s Watson, visit ibm.com/Watson.