January 20, 2017

Chinese robot gets byline in Guangzhou-based Southern Metropolis Daily


Bloop beep beep Russians boop blorp zoop Pizzagate zip zlorp tweet millions of votes

According to a report in the China Daily News, Xiao Nan, a robot designed and built by a team of researchers at Peking University, has written, delivered, and published a news article in the Guangzhou-based Southern Metropolis Daily. The 300-character story covers Chunyun, the Chinese spring festival that sees about 2.9 billion individuals returning to their home villages to celebrate the lunar new year.

Wan Xiaojun, who led the Peking University team, realizes that this kind of automation might, you know, freak journalists the fuck out, not least of all because “they fear they might lose their jobs.” But, by Wan’s own admission, “Robots are still unable to conduct face-to-face interviews and respond intuitively with follow-up questions. They also do not have the ability to select the news angle from an interview or conversation.” Which is to say that they are not competent journalists. Then again, neither are most human journalists. BOOM!

Xiao Nan’s newsroom invasion is not particularly new or groundbreaking, though it does continue a trend that’s been developing for a couple years. In June of 2015, The Verge ran a story on the Associated Press, which had started using news automation to generate stories on quarterly earnings reports. According to Automated Insights—the company that built the AP’s automation technology—the AP, using robotic reporting, “now produces nearly 3,700 quarterly earnings stories for US and Canadian companies, over 12 times the number that AP reporters and editors produced manually.”

Stories of robot reportage picked up steam in 2016. The Washington Post had bots on the Rio Olympics beat. British and Irish news agencies integrated robot writing into their sports and electoral coverage. And at The Atlantic, Adrienne Lafrance  published a piece about her experience training an AI to mimic her writing style. The result?

content that was that communications and everything that makes on a person what they’re are also to be in the Internet in the fact about it is that models are technologication of the same that its also from the most computer.”

A less than resounding success, I’d say.

The news industry is today facing a lot of really serious, systemic problems: how to optimize and stabilize a revenue stream that doesn’t make journalists wholly dependent on corporate largesse; how to successfully communicate with an increasingly polarized and isolated public; how to differentiate fact-based, objective journalism from the roiling sea of fake-ass tabloid news; how to do basic reporting on one of the most profoundly important news stories of the century. You know, big-league stuff.

The argument has been made that assigning robots to do the basic, repetitive gruntwork of sports reporting and financials frees up time for all the real journalists to real journalism. I’ll believe it when I see it.


Note: So, we’ve heard there’s some kind of big news story going on? There’ll surely be plenty to say about the American government in the weeks, months, and, ugh, years ahead. But, for today at least, we would prefer not to. Instead, we’re thinking about a few other things, and we invite you to think about them with us. Stay strong, listen to your heart, and we’ll see you on the other side.



Simon Reichley is the Director of Operations and Rights Manager at Melville House.