specifying peer-review or "official" context is a huge step IMO. hopefully gains like this can be made to keep AI "honest".

My wife is a 25+ year teacher. Like the unwanted invasion of Google products like 'Google Classroom' and cheap, constantly-breaking Chromebooks, teachers are generally horrified at 'products' like this entering public ed.

I can't tell you how much K-12 teachers generally want less tech in classrooms, not more. This benign little AI helper is just a trojan horse for another expensive line item for already strapped schools, and another distraction from the real challenges students face.