In what was probably an inevitable conclusion, X has that it will allow the AI of author of community notes. With a pilot program from today, the social network publishes developer tools to create editors of AI notes. These tools will be limited to the reward of responses in a test mode and will need approval before their notes can be released in the wild. The first authors of AI notes will be accepted later this month, that is to say at this moment that the compound notes will begin to appear to users.
“Not only does this have the potential to accelerate the speed and scale of community notes, but comments on community assessment can help develop AI agents that provide increasingly precise, less biased and largely useful information – a powerful feedback loop,” said the post announcing this feature.
It looks great. .
The editors of AI notes will be by “an assessor of automated notes with open source” which assesses whether the composition is on the subject and if it would be considered harassment or abuse. The decisions of the evaluator are based “on the historical contributors of the contributors of community notes”. Despite the insistence of the announcement of “humans still in charge”, it seems that the only human editorial eye comes from notes on notes.
Once the notes written by AI are active, they will be labeled as such as a measurement of transparency. The AI will only be authorized to propose notes on the articles which requested a community note at the beginning, but the company positions the authors of IA notes as having a more important future role in this system of verification of the facts.
