Why journalism schools are teaching students artificial intelligence
The Tow Center for Digital Journalism at Columbia Journalism School hosted an artificial intelligence and journalism workshop where particpants learned how to create AI-driven text and video stories.
As news consumption patterns change, reporters need to develop new storytelling approaches and master the use of new tools, many of which are now powered by artificial intelligence.
Implementing these smart machines in the newsroom will still demand editorial expertise, though. Because while content can be generated automatically, journalists in new roles such as “automation editors” will need to watch for and correct any errors.
“The current generation of successful automation technology scales the output of human labor, rather than replacing it entirely,” said Nick Haynes, a data scientist at Automated Insights who helped teach the workshop.
“These technologies have much more in common with power tools than sci-fi superhuman robots — in the hands of a skilled user, they act as a force multiplier in the speed, breadth and scale of content that can be produced.”
We opened the session at Columbia Journalism School with an overview of best practices when deploying new technology, followed by hands-on training on two tools widely used by newsrooms around the world: Automated Insights and Wibbitz.
Representatives from each platform gave students an overview of their software, answered questions and provided suggestions for how students could improve their automated content.
“Learning AI tools before hitting the newsroom will hopefully put me one step ahead in understanding how to save time on simple tasks and concentrate on more complex ones,” said Cecilia Butini, a Columbia Journalism School student who attended the workshop.
What the workshop facilitators found most surprising was how quickly every aspiring “automation editor” understood how to use these tools in such a short period of time.
“It was amazing to see a room full of future journalists dive in and start using a brand-new tool that they were just introduced to minutes prior,” said Zohar Dayan, the co-founder and CEO of Wibbitz. “Not only did they create great videos, but we also saw that they really enjoyed the process.”
Text Automation
During the workshop, participants learned how to turn data into a text report using Automated Insights, an AI tool that enables journalists to develop dynamic templates that convert structured data into human-readable articles. This approach works well for sports, finance and any other form of structured story
Video automation
Students learned how to operate Wibbitz, a platform that utilizes image recognition to create rough-cut videos composed of archived images and videos that match a given text automatically.
Here’s a great example of an automated video produced by two Columbia Journalism students, Deanna Paul and Taryana Odayar:
Artificial intelligence can help augment journalism, but it will never replace journalism. AI might aid in the reporting process, but journalists will always need to put the pieces together and construct a digestible, creative narrative.
“Pulling the veil of complexity back on computer automation allows journalists to not only publish more efficiently; it also augments our ability to report in an increasingly data-saturated and algorithmic world,” said Andrew Calderon, a graduate student at Columbia Journalism School who participated in the workshop.
