Image credit: artifact
Last month, the Artifact news app introduced the option for users to flag articles as clickbait. Now, the app, founded by Kevin Systrom and Mike Krieger, has launched a new feature where AI rewrites headlines if it finds such articles.
App makers say that when users mark a title as clickbait, the app calls the GPT-4 model to rewrite the headline.
At the time it launched its clickbait reporting feature, Artifact said it had a manual process to vet and mark articles reported by multiple users as clickbait.
With the new update, AI rewrites headlines for all users when an article is marked as clickbait by the Arificat team. A star icon appears next to the heading to indicate that it has been rewritten by AI.
Additionally, the startup is already working on technology to detect clickbait articles without relying on manual reporting. When this system is introduced, the app will automatically detect and rewrite headlines.
With these features, the app takes many editorial responsibilities on its own, such as correctly identifying clickbait articles and using AI to rewrite non-misleading headlines. However, the company does not hesitate to take such steps.
In an interview with TechCrunch in March, Systrom acknowledged that running a news app involves some form of editorial decision.
“In fact, building an algorithm is a very editorial task, because what you choose to train the algorithm—the objective function, the data to input, the data to include, and the data to exclude—is all editable. It’s a way of weighing different goals,” he said.