YouTube tests notes function to contextualise videos

The Google subsidiary YouTube is currently testing a new notes function. The test run is initially only taking place in the USA, as the company announced in its own blog. The function makes it possible to add notes to videos. These notes are intended to expand the context of the respective video, categorise the topic if necessary or provide additional information. The notes entered can also refer to specific points in time in the video. Users can also post links to other websites for further information.

Possible applications

The idea is not new. For example, X has already implemented such a Community Notes function. This is a tool that makes it possible to check the accuracy of articles on the platform and make corrections if necessary. This function has at least made a rudimentary contribution to curbing the spread of serious misinformation or identifying it as such.

Such references can not only serve to expose misinformation and put it into context. They can also serve as a service for further information: In documentaries, individual aspects can be provided with more in-depth references or links to other websites that cover the topic in more detail. Other videos, such as music videos or funny clips, can be exposed as prank. Or older film material that is falsely presented as current can be identified and labelled as such in this way.

How does the system work?

If notes are generally considered useful, they are presented publicly under a video. As part of the pilot project, feedback is obtained from external people. These are the same people who rate YouTube's search results and recommendations. They classify a rating as "helpful", "rather helpful" or "not helpful". The criteria for this include whether they are based on high-quality sources or whether they are written clearly and neutrally. In the further course of the process, the notes are also evaluated by the contributing creators.

The system then uses an algorithm to integrate these ratings and decide which notes are published. This algorithm is designed to help find notes that are useful to many people from different perspectives. If a note is rated as helpful by trustworthy people, it is more likely that YouTube's system will highlight this note under a video. People who have given different ratings in the past are considered trustworthy.

YouTube will also use the Community's ratings to train its own system. This should later take over the verification of the notes fully automatically.