Earlier this year, YouTube has brought a few updates to its platform this included a new feature called Chapters. This feature allowed creators to divide a video into separate chapters using timestamps. Chapters can be useful as it allows users to jump to the parts of the video that most interest them by clicking timestamps in the description or instead of scrubbing the video in the seek bar. That being said, not all videos uploaded comes with chapters as it involved the cumbersome task of manually identifying timestamps which can be a pain to do. To make things a little easier, YouTube has been testing an AI model which can divide videos into chapters on the fly.

To do this, YouTube is deploying AI that will go through a video and identify certain visual markers. These markers will then be used as reference points to break the video into chapters. The algorithm will also recognize certain text-based signs in a video to do the same. According to YouTube, the main purpose of this experiment is to create easy jumping on and off points for viewers, making it easier for viewers to navigate through videos and quickly jump to the relevant part that they desire.

The new feature is currently being tested on a small group of videos by YouTube. Needless to say, YouTube is allowing the creators to opt out from their experiment. That said, YouTube is also encouraging the uploaders to provide feedback on how the feature could be improved.