6 myths about robot journalism — busted
Big Data For News Publishers | 01 August 2021
Robot journalism as a concept has been around for some years now. The idea can inspire doubt and even dread in editors and journalists: Will the robots replace reporters? Who controls them and how? Can they be trusted?
At United Robots, we and our publisher partners have worked with the reality of news automation since 2015. We see how robot-written content actually supports journalism, and we think it’s time to address some of the popular myths that surround it.
Myth 1: Robots steal journalists’ jobs
In our experience, this is not true. What they do is free up journalists’ time by doing the routine, repetitive reporting that, quite frankly, journalists are overqualified to do. Robots also allow publishers to cover huge volumes of hyper-local stories. This is coverage that would be completely unfeasible to do manually — like write about 60,000 local football matches, as is the case at Dutch local media group NDC.
Much has been written about how robots will allow news publishers to cut costs in the newsroom. Among the publishers we work with, this has never been the motivation behind automation. In fact, the local media groups we collaborate with in Scandinavia, the Netherlands, and the United States, for example, all have a strategy of improving and investing in local journalism to drive loyalty with local readers. Robots help them expand their hyper-local coverage and eliminate news deserts. In fact, many of them are actually recruiting journalists as well as deploying robots.
Myth 2: Robots use repetitive language
Yes, but that’s not the whole story. Robots should do the routine, repetitive reporting. Journalists do the human stories. And while our robots do volumes of routine stories, we build them for article and language variation and to each publisher’s particular editorial standards.
Myth 3: Robots can’t be trusted
Actually, robots make fewer mistakes than humans. The technology United Robots uses creates texts from structured data. If a particular fact is included in the data, it will be in the story. If it isn’t there, it won’t be in the text — and it certainly won’t be replaced with something else. Our publisher partners trust the robot content. The vast majority of it is published straight to their readers.
Myth 4: Robots’ texts need manual checking
The structure of the texts and rules around what angles the robot looks for are defined in detail by the newsroom in collaboration with us. Since the texts are based on structured quality data and the robots follow their predefined rules, the results are texts that are safe to publish automatically and directly to end users.
Myth 5: Robot journalism is technically difficult
Well, it is if you’re going to build your own robot. But publishers can use a third-party partner to build the robot for them. We help the newsroom train the robot to write to its particular editorial style and guidelines. And once that’s done, the automated content generation begins. We look after the robot, and the publisher gets their texts and images sent directly to the CMS, sites, or apps.
Myth 6: To have control of the robots, we need to build them ourselves
Actually, it can be risky to rely on internal experts. Working with a company like United Robots removes that risk. Our robots continue to work irrespective of tech experts coming and going. And the control is with the publisher. We help newsrooms set the framework and rules that the robots work from.
Robot journalism may be a new and relatively untested newsroom resource. However, the learnings and best practices that do exist suggest that rather than being a threat, done right it actually provides opportunities — especially for local journalism.