Moderation Process

Every contribution you submit is evaluated by an automated process before it can enter the catalog of published study content. Once you submit your file, the system extracts its content. The text is read page by page, and then it identifies and processes any tables, as well as images inside it. For images, and for scanned files, the content is read by AI vision after it is classified.

The extracted content now becomes a part of the AI Teachers' Council. It is a council of 5 AI teachers who specialize in their particular area of expertise. Luna the Creator reads the original content and creates the first draft of study content for it. Atlas the Reviewer reads the content, and evaluates its factual accuracy, comprehensiveness, and educational quality. Nova the Enricher adds more detail and examples, improving on the clarity and depth, and polishing the language.

Sage the Validator reads the content, and evaluates whether the factual claims in the content are consistent and accurate, with respect to the original source content. Echo the Fact Checker evaluates and crosschecks the factual claims in the content, against factual claims about the world. No teacher reads the work of others before making a judgment on it themselves. If all 5 teachers agree, the content is published. If any teacher identifies problems, the contribution is rejected. Rejected contributions are sent a message with a reason for the rejection, and you can choose to appeal the decision. All of these steps together are how we ensure that only high quality and accurate material is made available to other students.