AI is now part of the edtech stack — and schools are repeating the mistake they made in every wave before

Your company just activated the most capable technology it has ever bought.

The AI assistant arrived the way the 18 systems before it arrived — the CRM, the HRIS, the data platform, the two collaboration suites that do the same thing — with a launch email, a recorded webinar you attended in spirit, and a dashboard showing near-total activation by Friday.

Ask what it has changed about the output and you get a shrug and a workaround somebody built in a spreadsheet.

Activation is magnificent and output is flat, and every executive reading this has sat in that meeting.

MIT put a number on it last year: 95% of corporate generative AI pilots produced no measurable return.

American schools just ran the same play, at national scale, and are about to be graded on it.

The interactive whiteboards, the laptop carts, the platforms bought with pandemic money — each one activated without the training and instructional model built around it.

There was always an alibi — the tool was clunky, the migration was botched.

AI takes the alibi away.

This time the technology is genuinely excellent, and the results still are not moving.

That settles an argument the software industry has dodged for a decade: the constraint was never the tool.

It is everything nobody builds around it — the training, the workflow, the taught habit of using it well.

AI makes the shift from adoption to outcomes undeniable, and Washington last week started asking schools for proof.

Coming out of the pandemic, I watched a version of this play out in our schools as they tried to implement a new wave of career exploration beyond the standardized test and the familiar conversation with a guidance counselor.