I'm a serial edtech founder, currently building a product based on generative AI (like everyone else in my sector). The stumbling block for me has been quality questions - AI is capable of creating and marking basic questions like adding fractions or asking the reasons the enlightenment happened, but in my experience it fails when it comes to creating structured, well thought out questions in which one part develops from another, such as those that appear in A Level or University level questions.<p>In looking for such questions, it surprises me that there is no open question bank, filled with quality questions that can be used by teachers, students and educational businesses/charities. It would lower the barrier to entry for edtech entrepreneurs substantially.<p>I have the ear of some big charitable funders, and am thinking of proposing the creation of such a bank with:<p>- a range of high quality questions across all subjects and topics
- matching mark schemes, built and tested with a range of LLMs for providing automarking and advice to students<p>While some of the questions would go out of date, and there would likely need to be some ongoing pruning, I suspect such a bank continue to provide value for many years.<p>As a long time daily reader of HN, I would very much value your thoughts on whether this is a good idea or not, and how it could be made the most useful.<p>Thanks HN!