Dan Meyer

Personalized learning has been the dream of edtech for the last idk 100 years, since at least programmed radio instruction. The mistake boosters keep making is thinking the problems are technical and solvable...the bigger issue is that fundamentally most people do not want this.

Serious question. One of the oft theorized use cases for AI is as a personal tutor. Sounds compelling, but isn’t this incredibly far off? We have computers that can destroy a human at chess. Yet none, AFAIK, that can explain its moves to a student.

Still blows me away that memorization somehow became taboo in so many classrooms. Memory is a precondition for thought, and thought is the precondition for learning.

I agree with you, but I think it's appropriate to drag efforts which make extreme claims while either naively or deliberately failing to acknowledge/compensate for selection effects. It's a vibe thing for me: just feels unserious, epistemically. It's not surprising to me that high-performing students learn faster at a school with only high-performing students. That's consistent with pre-digital "gifted" programs; AFAICT those historical results are mostly a matter of permitting a faster pace, rather than "advanced technologies" (which appear to be CAI designs from the 80's?). What bothers me is that their marketing is just, flatly, "Alpha students learn twice as fast as their peers and rank in the top 1% nationwide". Which—okay, sure, but the implication of that language is "Alpha can cause a median student to learn twice as fast". The data don't support that; the language is misleading. More rant from me in a group chat on Alpha a while back: