My journalism career began when I joined the staff of my high school newspaper during freshman year. Three years later, I was named sports editor - a curious choice, since I neither knew nor cared much about sports.
Never one to be deterred by ignorance, I prepared for basketball season by taking the sports sections of Tampa's two daily newspapers and jotting down samples of basketball lingo. After perusing several articles, I knew two dozen ways to describe the act of making a basket.
It took me a lot longer to understand that good writing required more than stringing clichés together.
A lifetime later, I now realize I could have been replaced by a computer. The New York Times reported recently that artificial intelligence is being used to compose news stories.
Before anyone challenges the premise that journalism and intelligence should be linked in the same sentence, let me elaborate:
According to The Times, a firm in Evanston, Ill., has created a program that takes data, such as sports statistics or building permits, and turns them into readable articles. The technology has been employed by the Big Ten Conference and Fox Networks for short summaries of baseball and softball games since last year.
For years, the article explained, programmers had tried to create software for basic articles, but the results tended to follow a fill-in-the-blank format and read as if they had been written by R2-D2.
After a decade of experimentation, researchers have come up with software that can make inferences based on historical data, according to The Times. The computer can even develop "story angles" by learning concepts such as "individual effort," "team effort," "come from behind," "back and forth," "season high," "player's streak" and "rankings for team." The computer analyzes what element is most important for a given game, and that becomes the lead of the article, according to one of the software developers.
Advocates of this trend argue that the articles created by "robot journalists" replace low-level coverage that many newspapers no longer can afford.
Granted, loss of subscription and advertising revenue to the Internet has caused newspapers in this country to reduce their "news hole" and to slash newsroom jobs - salaries and newsprint being the industry's biggest expenses.
Knowing that computers can defeat international chess masters and wipe the floor with former "Jeopardy" champions, should we be surprised that a computer could write a news article?
Most reporters will admit that they could put their brain on autopilot and still write a credible story about a routine city council meeting or your garden-variety bank heist. There are just so many ways to describe squabbles over sewer fees, a Saturday-night stabbing or the daily carnage on our highways.
Nor would the content of your typical TV news program present a significant challenge to robot journalism. Let me put it this way: Once geeks figure out how to waterproof a computer, The Weather Channel's Jim Cantore could be looking for a job.
Ironically, Narrative Services, the company that developed the robo-news technology, is based at Northwestern University, home to one of America's most respected journalism schools.
Arguments being advanced in support of robot journalism sound suspiciously like the rationale spouted in defense of outsourcing jobs to China or India. Why not let a computer produce no-brainer stories, so that journalists can concentrate on the important stuff?
Right! Bean counters calling the shots for publicly traded media companies would take any money saved by employing robot reporters to the bottom line. What's a Pulitzer Prize compared to the price of company stock?
Unfortunately, no one seems to acknowledge that the articles most ripe for plucking by robots are the very stories that formerly were assigned to rookie reporters.
Many a good journalist cut his or her teeth writing obituaries or taking game stories from high school football coaches over the phone.
In the process, they learned basics of news writing, the need for accuracy and just what was important to the community they served.
Try teaching that to a computer.
Editor's note: This column was composed by a human being.
Comments