4 ms·
> split the day between teaching and learning (probably alternating by hour) In peer instruction[1], the lecturer asks a question, students each commit to an a
by mncharity 8y ago
> split the day between teaching and learning (probably alternating by hour)
In peer instruction[1], the lecturer asks a question, students each commit to an answer (eg using clickers, or held-up cards), then turn to discuss the question with peers, and then do another commit. In computer-supported peer instruction, when students turn to discuss the question with peers, an app tells each student which of their neighboring students to talk with, based on the students' recent clicker answers, to maximize productive discussion.
So for the future, perhaps we could do better than dividing teaching from learning by hour or day or year; better than partitioning teachers from learners by grade. Perhaps instead imagine dynamic minute-by-minute formation and dissolution of variously-sized groups for optimized and personalized learning.
AI: "Hmm, student A's puzzlement will soon exceed their frustration target profile, and student B just figured it out and is enthused, and student A-B pairings have a history of going well in this situation... message: 'A,B: I suggest you pair'".
[1] https://en.wikipedia.org/wiki/Peer_instruction https://en.wikipedia.org/wiki/Peer_instruction
- sitkack 8y agoCould you expound on this more generally? I think it could be excellent prior art for pairing student/student and student/teacher.
- mncharity 8y agoI'm unclear what you're asking for? The computer-supported peer-instruction phone app example is based on work by Eric Mazur at Harvard (physics; physics education research; peer instruction). The AI example is simply a generic computer-human process hybrid, taking the existing teacher/tutor task of productively managing student grouping and learning, and scaling it with computer-based monitoring and optimization. Prior art? If your interest is in patents, my fuzzy understanding is Mazur regrettably took a familiar path. Unable to get follow-on research funding on the topic, he created a startup, with Harvard patent(s) (on what exactly I must of course avoid knowing) for table stakes and moat. The startup of course failed, and was acquired by a large dysfunctional education incumbent for the anticompetitive "value" of the patents. Leaving society arguably much worse off than if the original physics education research had never been funded at all. It's my fear that the conjunction of education with computer-human hybrid processes, and with with AR/VR, given the regulatory capture of patent policy, and plummeting ordinary skill levels, is creating a toxic stew of patents that will profoundly impede innovation in education in the US for many years.
- sitkack 8y agoThe https://en.m.wikipedia.org/wiki/Bayh–Dole_Act https://en.m.wikipedia.org/wiki/Bayh–Dole_Act strikes again! Thank you.