Did a robot write this? To many, such a question sounds absurd. But thanks to artificial intelligence (AI), specifically the development of machine learning, it’s becoming more of a reality with each day that passes.
Welcome to the new era of AI-assisted journalism.
A Race to Break News
Approximately one-third of the content from Bloomberg News leverages one form of automation or another. Thanks to a system called Cyborg, the publisher’s reporters are able to produce thousands of articles about company earnings each quarter.
With no need for breaks, lightning-fast comprehension, and pinpoint precision, Cyborg can digest a financial report and return a news story with the most relevant facts and figures in a matter of minutes. It allows Bloomberg News to compete with rivals such as Reuters in the fast-paced world of financial journalism.
Finance has always moved at a rapid-fire pace; tomorrow’s insights often become yesterday’s news in only a few moments. But other news outlets have taken note of robot reporters’ untiring accuracy and unparalleled efficiency. Today, you can find AI helping to break news on earthquakes for the Los Angeles Times and giving play-by-play game reviews of minor league baseball for the Associated Press.
With that being said, is this the end of human journalism?
A Threat? No. A Tool? Yes.
Practically every major publication has its eyes on AI now. A few weeks ago, The Guardian’s Australia branch published its first AI-assisted article. And Forbes recently announced it would be testing an AI to help reporters with rough drafts.
It’s easy to see that AI is quickly becoming part of every journalist’s toolbox. But industry insiders are quick to point out that it’s not a threat to their jobs. Instead, according to Lisa Gibbs, director of news partnerships for the AP, AI will enable journalists to spend more time on more important work:
“The work of journalism is creative, it’s about curiosity, it’s about storytelling, it’s about digging and holding governments accountable, it’s critical thinking, it’s judgment — and that is where we want our journalists spending their energy,” explains Gibbs.
Early Adoptions = Early Wins
The AP is one of the first adopters of AI. In 2014, the publisher began working with Automated Insights, a tech company that utilizes natural language generation software to produce billions of narratives per year.
At first, the software was mostly used to cover college sports news. Then, like Bloomberg News, the AP began expanding its coverage of company earnings. Before Automated Insights, the news agency would produce around 300 earnings report articles per quarter. Today, it churns out 3,700 per quarter.
The Washington Post is another early adopter of AI. In 2016, it used Heliograf, its in-house robot reporter, to cover that year’s Summer Olympics and the presidential election. Thanks to Heliograf, the newspaper took home the “Excellence in Use of Bots” award at the Global Biggies Awards last year. This annual ceremony recognizes achievements in big data and AI.
The Nuts ‘n’ Bolts Behind the Story
Before you get the wrong idea, it’s important to emphasize that human touch is still a huge factor in many of these AI-assisted stories.
Bloomberg News, the Washington Post, and the AP all utilize internal alerts to detect and indicate instances of irregular data. Reporters then look at this alert and try to determine if it’s a story that should be handled by a human or not. For example, during the 2016 Olympics, the Washington Post had thresholds set to alert editors if an event result was either 10% above or below the current world record.
Even when a robot is handling the story, it’s not as easy as giving them a typewriter or keyboard. As far as the front end goes, many of these publications still have writers craft a story for each possible outcome of an ongoing event. Once the final results come in, the AI then creates the article.
Just like human-generated journalism, AI-crafted stories are far from infallible. For example, many companies have begun to carefully choose the figures released for their earnings reports. They know that certain information will result in a more favorable depiction. Bloomberg News is already aware of this and taking action; the publisher is preparing Cyborg to avoid such manipulation.
The Future of Information Exchange
We’re at the beginning of a new era in journalism, one in which AI is quickly becoming an indispensable tool. And not just in the case of writing rote articles.
“I hope we’ll see A.I. tools become a productivity tool in the practice of reporting and finding clues,” says Hilary Mason. Mason is the general manager for machine learning at Cloudera, a San Francisco-based data management company. “When you do data analysis, you can see anomalies and patterns using A.I. And a human journalist is the right person to understand and figure out.”
Both Dow Jones and the Wall Street Journal are already trying out the technology to take on a variety of tasks such as transcription or identifying deep fakes. Francesco Marconi, head of R&D at the Wall Street Journal, says “Maybe a few years ago A.I. was this new shiny technology used by high tech companies, but now it’s actually becoming a necessity. I think a lot of the tools in journalism will soon be powered by artificial intelligence.”
Above all else, it’s important to remember that change is the only constant. And staying adaptable is the only way to stay ahead of the curve. Marconi sees the introduction of AI in newsrooms as comparable to when telephones first showed up: “It gives you more access, and you get more information quicker,” he said. “It’s a new field, but technology changes. Today it’s A.I., tomorrow it’s blockchain, and in 10 years it will be something else. What does not change is the journalistic standard.”
What do you think of AI entering the news arena? Do you think it will result in better, more informed stories? Let us know your thoughts in the comments!Tags: AI and machine learning, AI and ML, AutoML, benefits of machine learning, machine learning, machine learning app developer, machine learning app developer San Francisco, machine learning app development, machine learning app development San Francisco, machine learning applications, machine learning apps, machine learning developer San Francisco, ML, ML and AI, San Francisco machine learning app developer, San Francisco machine learning app development