The platform announced the integration of the Pangram program, which specializes in detecting materials written using artificial intelligence tools, to allow users to examine the posts, comments and replies on the platform to determine how much of the material was written by the author himself and how much of any publication was written with artificial intelligence tools.
The website TechCrunch, which specializes in technology topics, stated that this feature could harm Substack’s activity in the short term, because it could reveal that many of the news bulletins on the platform are not entirely original. This could also reduce the platform’s history in the system as an independent source of news and blogs, and even negatively affect its reputation as a platform for hosting high-quality content.
But in the long term, AI usage detection features could help the platform weed out “bad AI content” and encourage more users to trust that what they read on the platform is the original production of participants, or at least know it isn’t.
With this step, Substack joins many content publishing platforms in various forms that provide tools that help users distinguish between content produced by real people and content generated by artificial intelligence, whether in whole or in part.
Social media platforms are already tagging photos and videos generated with AI, while many music streaming platforms have started tagging tracks composed using AI, and in some cases deleting them entirely.
For his part, Chris Best, CEO of Substack, says: “This is a good use of artificial intelligence… When I introduced the Substack platform to writers, one of my approaches was to say that we will do everything for you except the hardest part. You have to have something – an idea that’s worth reading, that’s worth paying attention to, that’s worth sharing that part… that’s really hard to do that part… and the platform does everything else” for the content to reach the audience.
