What the world's most respected education thinkers are saying about AI
Nov 11
0 comments

What the world's most respected education thinkers are saying about AI

Four Brilliant Minds. One Uncomfortable Truth.
What the world's most respected education thinkers are saying about AI — and what it means for how you teach.

Over the past few months, I've been reading everything I could find from the world's most respected education thinkers — not what they wrote ten years ago, but what they're saying right now, post-AI.

Howard Gardner at Harvard. Mitchel Resnick at MIT. Yong Zhao at the University of Kansas. And economist Daniel Susskind, whose new book asks the most honest question a parent can ask: what should my children actually do?

They come from different disciplines. They write in different styles. They arrive at the same uncomfortable place.

Howard Gardner
Harvard University · Psychologist · Creator of Multiple Intelligences Theory

Gardner says AI will soon outperform even the most gifted human teacher at instruction. His conclusion isn't despair — it's clarity. Two minds cannot be downloaded onto a device: the respectful mind (genuine care for other people) and the ethical mind (the ability to navigate hard moral questions as a citizen and professional). These are what school must protect.

"Well-designed AI, I believe, will outperform even the most gifted instructor." - Five Minds for the Future (revised 2024) · Who Owns Intelligence? (2024) · Education Week op-ed (2026)

Mitchel Resnick
MIT Media Lab · Lifelong Kindergarten Group · Creator of Scratch

Resnick warns that AI will be deployed in classrooms in the most passive way possible — as a smarter worksheet machine. What he advocates instead: the 4Ps (Projects, Passion, Peers, Play) as the conditions under which children become Creative, Curious, Caring, and Collaborative. His point isn't technological. It's political.

"It is up to us — the choice is educational and political, not technological." - Generative AI and Creative Learning (MIT Press, 2023) · ISTE 2025 keynote · Human-Centred Society essay (2025)

Yong Zhao
University of Kansas · Education Policy Reformer

Zhao says AI hasn't broken education — it's made the cracks visible. He calls it the "grammar of schooling": same age, same curriculum, same test, same pace, same outcome. His prescription: give students 40% self-directed time. Shift from problem-solving to problem-finding. Trust the kids.

"AI did not create this obsolescence. It revealed it." - Artificial Intelligence and Education: End the Grammar of Schooling (2024) · Fix the Past or Invent the Future (2025)

Daniel Susskind
Oxford University · Economist · Author

Susskind puts it in terms parents understand: don't train for a job title — train for the problem. Build foundations, not forecasts. Teach with AI and without it, the same way a 1982 math teacher taught with a calculator and without one. And consider learning a trade — AI still struggles most with physical, hands-on work.
"Don't train for a job title. Train for the problem." - What Should My Children Do? How to Flourish in an Age of AI (2026)


I'll be honest. AI in education doesn't scare me. What scares me is using AI to do more of the same — faster.

After reading all four of these thinkers, the thing that keeps staying with me is something Resnick and Zhao both circle around: not every child needs to reach the same place. What matters is that each child's own capacity gets pushed — in the direction that is theirs. Their curiosity. Their way of thinking. Their particular strength.

That is what play-based learning has always done well. It doesn't deliver answers — it creates conditions. A child spinning a story wheel doesn't know what story will come out. A child sorting phonics tiles doesn't know which word they'll find. That not-knowing is not a gap in the design. That's the design.

The human act of generating something new from your own mind — that's the one thing AI genuinely cannot replicate. And it turns out, it's also the thing most classrooms have slowly been designing out.

So this issue is for every educator who already feels this — but hasn't had the language for it yet. You do now.

  1. Try the problem-finding flip
    In one lesson where you normally ask students to solve a problem — give them the context instead, and ask them to find the question. See what happens.
  2. Look at your assignments through Zhao's lens
    Could a student complete this using only AI? If yes — that's not a cheating problem. That's a design problem.
  3. Ask your school one question
    "What percentage of a student's week is self-directed?" If the answer is zero, Zhao would say you're running a 19th-century school with 21st-century branding.
Free Resource
Is your school ready for the age of AI?

A free 10-question audit for school leaders and educators. Get an honest picture of where your school stands — across curriculum design, assessment, teacher role, student agency, and learning environment.

Take the School Audit →

Takes 10 minutes. No rankings. Just clarity on where to move next.

Leave a Reply