A year after this talk, my Google-sponsored, Futures Lab cohort at UW was challenged to "use AI", meaning "come up with a prototype tool using Gemini Pro to vibe code the demo", to create a learning tool. Without even really directly sourcing from this video - all 4 on my team snapped our focus on to the problem of creating with so-called "AI" accidentally displacing the appropriate and necessary *effort* that is actually what makes learning happen. I'd not consciously realized the connection happened until near the end of our project.