Your newest Intelligencer blogger.
Earlier this week, one of my business-beat colleagues got assigned to recap the quarterly earnings of Alcoa, the giant metals company, for the Associated Press. The reporter’s story began: “Alcoa Inc. (AA) on Tuesday reported a second-quarter profit of $138 million, reversing a year-ago loss, and the results beat analysts’ expectation. The company reported strong results in its engineered-products business, which makes parts for industrial customers, while looking to cut costs in its aluminum-smelting segment.”
It may not have been the most artful start to a story, but it got the point across, with just enough background information for a casual reader to make sense of it. Not bad. The most impressive part, though, was how long the story took to produce: less than a second.
That impossible-sounding deadline was possible because the AP’s story wasn’t written by a person at all. It was the product of a piece of software — a robot, really — created by a Durham, North Carolina-based company called Automated Insights. The AP announced last month that it would use Automated Insights’ software, called Wordsmith, to produce up to 4,440 robot-written corporate-earnings reports per quarter, more than ten times the number its human reporters currently produce. When Alcoa’s earnings report hit the wire that day, the data was instantly compiled by a firm called Zacks Investment Research and passed through the AP’s proprietary algorithm, which pulled out key numbers and phrases and matched them against other contextual information. In milliseconds, the software produced a full story written in AP style, indistinguishable from one made with human hands. In short order, the story appeared on many AP member sites, including NPR, Hawaii News Now, and KAIT8. (According to an Automated Insights spokesman, human AP employees were involved in editing and adding to the Alcoa story after the fact, but the AP has said that in the future, the process for many stories will be fully automated.)
By this point, we’re no longer surprised when machines replace human workers in auto factories or electronics-manufacturing plants. That’s the norm. But we hoity-toity journalists had long assumed that our jobs were safe from automation. (We’re knowledge workers, after all.) So when the AP announced its new automated workforce, you could hear the panic spread to old-line news desks across the nation. Unplug the printers, Bob! The robots are coming!
I’m not an alarmist, though. In fact, I welcome our new robot colleagues. Not only am I not scared of losing my job to a piece of software, I think the introduction of automated reporting is the best thing to happen to journalists in a long time.
For one thing, humans still have the talent edge. At the moment, the software created by Automated Insights is only capable of generating certain types of news stories — namely, short stories that use structured data as an input, and whose output follows a regular pattern. The software’s algorithms comb data feeds for facts and key trends, and combine them with historical data and other contextual information to form narrative sentences. Currently, Automated Insight’s software is being used to write fantasy-football recaps for Yahoo Sports and internal sales reports for Fortune 500 companies, create real-estate listings, and do other, similar types of simple reporting. The Los Angeles Times has even built automated earthquake-reporting software, which was put to the test earlier this year.
Robot-generated stories aren’t all fill-in-the-blank jobs; the more advanced algorithms use things like perspective, tone, and humor to tailor a story to its audience. A robot recapping a basketball game, for example, might be able to produce two versions of a story using the same data: one upbeat story that reads as if a fan of the winning team had written it; and another glum version written from the loser’s perspective.
But these robots, as sophisticated as they are, can’t approach the full creativity of a human writer. They can’t contextualize Emmy snubs like Matt Zoller Seitz, assail opponents of Obamacare like Jonathan Chait, or collect summer-camp sex stories like Maureen O’Connor. My colleagues’ jobs (and mine, knock wood) are too complex for today’s artificial intelligence to handle; they require human skills like picking up the phone, piecing together data points from multiple sources, and drawing original, evidence-based conclusions.
The stories that today’s robots can write are, frankly, the kinds of stories that humans hate writing anyway. I should know; at my last job at the New York Times, I was occasionally called on to write quick recaps of corporate earnings, much like the ones the AP’s robots now write. It was a miserable early-morning task that consisted of pulling numbers off a press release, copying them into a pre-written outline, affixing a headline, and publishing as quickly as possible so that traders would know whether to buy or sell. The stories were inevitably excruciatingly dull, and frankly, a robot probably could have done better with them. (The bot could at least have given some spark to a sentence like “BlackRock, the asset management giant, reported a fourth-quarter profit of $555 million, down 16 percent from a year ago.”)
If the Times had used an earnings-story robot rather than passing the task to cub reporters, I could have spent my mornings working on stories that required actual human intelligence — profiles, trend pieces, voice-driven analysis. This is exactly why I’m not worried about the coming of the robot-reporters: rather than putting us out of work, it might free us up to do more of the kinds of work we actually like.
Robbie Allen, the CEO of Automated Insights and a former Cisco engineer, agrees. “The publishing industry at large is facing lots of challenges with its business model, and new things have to be tried out,” he told me in an interview this week. But he insisted that “new things” doesn’t just mean mass layoffs. “I’m not worried for reporters at all,” he said. “Our technology simulates what a data analyst does. It doesn’t simulate what a reporter does all that well.”
What a robot can do, Allen said, is churn out stories at a superhuman pace. Last year, Wordsmith, the company’s software, produced 300 million stories — more than every other media outlet in the world combined. This year, Wordsmith is expected to work even harder — producing more than a billion stories. Few of these will be gripping reads. Most will be sports recaps, personalized financial reports, or other commodified, just-the-facts types of news. But together, they’ll make up the biggest feat of news production in the history of the world. More isn’t necessarily better, of course, but there’s something to knowing that whatever obscure sports team you’re following, whichever esoteric stock you’re curious about, there will be a full, readable report waiting for you, instantly, at every turn.
“Most of our implementations have not been replacing what already exists,” Allen said. “You can’t find reporters who are are willing to write ten thousand fantasy football recaps a week.”
Ken Doctor, an independent analyst who studies the news industry, told me this week that the rise of robot reporting is a product of the times — both technologically and in terms of the troubled economics of the news industry, which has led media organizations to search for ways to reduce their costs. “The robots are just another tool of new journalism,” he said. In the future, Doctor predicts, robots won’t just be reporters’ competitors. They’ll collaborate with us by preparing data-dense paragraphs that we can then supplement with our own analysis, producing a hybrid story that’s better than our human efforts alone.
“Journalism is becoming a more highly skilled job,” Doctor said. “Simply showing up, in the Woody Allen sense — being able to read a press release or interview a single person, and write up a story that is understandable in 750 words — that’s not going to be enough. The optimistic part of this is that we’d use computers to do the basic work of organizing facts, and that the judgment and analysis, the interpretation, the experience is brought to it by humans.”
Robot assistance may even spur human reporters to do our jobs better. With software producing the equivalent of old-school “clip files” for us, we’ll essentially have full-time research assistants. The information in our stories will be more accurate, since it will come directly from data feeds and not from human copying and pasting, and we’ll have to issue fewer corrections for messing things up. Plus, with our nuts-and-bolts reporting out of the way, we’ll be able to focus on the kinds of stories that educate and entertain readers in a deep way, rather than just dragging simple information from Point A to Point B. Perhaps we’ll even start sprucing up our basic news stories, just to prove to readers that they’re not written with automated help. (“Could a robot have made a Beyoncé reference like this?”)
I’m not trying to be overly flip. The world needs earnings reports and sports recaps, and right now, it’s someone’s job to write them. The automation of news will, in all certainty, reduce the need for human reporters, especially at the local and regional levels, where funding is sparse. And as the technology behind content-creating robots gets more sophisticated, their reach will extend to other types of news. TV recaps, aggregated blog posts, sports analysis, political campaign coverage, BuzzFeed listicles — all of these could, conceivably, be automated, and the humans currently producing them could be cut out of the picture. The robots could even make inroads into PR and communications; according to Allen, several political organizers, including some former Obama campaign staffers, have already reached out to Automated Insights about using Wordsmith to create personalized voter outreach emails.
But even in the worst-case scenario, the rise of automated reporting is much, much less dangerous for the career prospects of journalists than the basic economic difficulties that are already facing the news business — the migration of print content online, the dry-up of advertising revenue, et cetera. That’s reassuring on some level, and terrifying on another. But the upshot is that the robot revolution isn’t something for most writers to fear. If anything, it will only make our jobs easier and more fun. We should be celebrating the rise of machines that can supplement and assist us in our jobs, while doing the most banal parts of our workloads for us.
After all, if we wanted to do something mindless and repetitive with our lives, we’d have become robots.