Memorex had a series of TV commercial in the 1970s that challenged the viewer to determine whether what they were watching was live or whether it was a recording, captured on Memorex tape.
Those commercials came back to me today while reading a story about robo-journalism.
Consider the opening sentences of these two sports pieces:
“Things looked bleak for the Angels when they trailed by two runs in the ninth inning, but Los Angeles recovered thanks to a key single from Vladimir Guerrero to pull out a 7-6 victory over the Boston Red Sox at Fenway Park on Sunday.”
“The University of Michigan baseball team used a four-run fifth inning to salvage the final game in its three-game weekend series with Iowa, winning 7-5 on Saturday afternoon (April 24) at the Wilpon Baseball Complex, home of historic Ray Fisher Stadium.”
One of those sentences as written by a human, the other by a computer program (*answer at end of post).
Apparently the use of computers to generate news stories has been in use for a while, and the article refers to this as “automated narrative generation.”
Using algorithms and natural language generators to create written content has been in use for a while; the article refers to this as “automated narrative generation.” The programs just take data, such as sports statistics or financial data, run it through its algorithms, and out pops a story.
And these algorithms don’t just spit the data back to the reader in a way that would make it obvious that a machine is behind the whole charade. These programs can write the stories in any voice, from serious to snarky, and for any audience.
The article claims that there is a need for such programs, given how much data we are inundated with on a daily basis. There aren’t enough people to process and report on all that data, hence the perceived need for such programs.
The Associated Press uses Automated Insights’ Wordsmith platform to create more than 3,000 financial reports per quarter while Forbes uses Narrative Science’s Quill platform for similar efforts. The Los Angeles Times using an algorithm called Quakebot to analyze geological data, was the “author” of the first news report of the 4.7 magnitude earthquake that hit Southern California last year, published on the newspaper’s website just moments after the event.
While the claim is that robo-journalism will free humans to do more reporting and less data processing, Kristian Hammond, Narrative Science’s co-founder, estimates that 90 percent of news could be algorithmically generated by the mid-2020s, much of it without human intervention.
And it’s not just the news that this has impacted.
Philip M. Parker, a management science professor at the French business school Insead has a patented algorithmic system which has generated more than a million books, more than 100,000 of which are available on Amazon. Give him a technical or arcane subject and his system will mine data and write a book or report on that topic.
These robo-journalists raise some issues.
First, the algorithms are missing the attributes that make us human, such as empathy, insight, and curiosity.
Second, it’s possible that the algorithms have biases built into them that are not always obvious.
This reminds me of the story that came out yesterday about Facebook. According to unnamed former contractors who curated for Facebook’s Trending section — a section that’s separate from users’ main news feed — news from conservative publications was suppressed by other curators, and that some topics were “injected” into the section even though they weren’t trending.
The Facebook executive who’s in charge of the social network’s Trending Topics said today that the company has “found no evidence that the anonymous allegations are true.”
What’s interesting is that it wasn’t a computer algorithm deciding which stories to include in the trending section, but humans. But the outcome may be a move on Facebook’s part to begin using algorithms to pick its Trending stories as a way of avoiding such accusations in the future.
I’m not so sure such an algorithmic solution would alleviate such concerns, since the algorithms are created by humans, who have biases. How can we be sure such biases wouldn’t work their way into the computer programs?
But my concerns don’t mean that I don’t expect a continued increase in the use of such algorithms; I’m sure we will.
It’s just that the question for today is not whether something is it live or is it Memorex, but whether it’s human or robot.
*the first story was created by a robot, the second by a human. If you would like to further test your ability to discern such differences, there is an interactive quiz on the New York Times web site. And if you would like to read an interesting article about the potential problems associated with robo-journalism, here ya go.
**picture from Muck Rack
P.S. Despite the lack of emotional depth exhibited in this post, you can rest assured it was written by a human…