I think both things can be true at the same time. There’s a bubble behavior around AI (valuation, marketing, thin wrappers), but the underlying capability shift is real. In education specifically, the most tangible value I see so far is on the productivity side (teacher leverage, content creation, admin) and structured student practice, rather than fully autonomous teaching. The challenge is that many tools look impressive in demos but struggle to prove durable learning outcomes or ROI, especially given slow procurement cycles. My takeaway is that AI in education is likely transformative, the winners will be the ones embedded in workflow, tied to measurable outcomes, and built with real pedagogy and trust.

This is best description of what's happening in AI that I've heard. Tim Ferriss and Bill Gurley discuss how every real technological innovation comes with speculative bubble behavior as well. redacted