Education in an AI-Shaped Future

Education cannot wait for certainty.

AI is changing how people access information, perform cognitive work, and think about expertise. Its longer-term effects remain uncertain, but our education systems cannot wait for complete certainty before asking what may need to fundamentally change.

This working group will bring together a small number of people to explore the questions carefully. In practice that means curiosity and intellectual humility: separating what we know from what we are guessing, and being specific enough to be wrong.

Questions we will explore

One question per week, in this order.

Who it is for

This group is for people thinking seriously about education, learning, AI, policy, pedagogy, or adjacent areas.

You do not need a formal background in education or technology. Curiosity, openness, and a willingness to engage thoughtfully matter more.

Where it is anchored

Anyone may apply, wherever they are. Three of the four sessions are not specific to any one country; the fourth is entirely about India. Sessions run at midday IST, which sits comfortably across Asia, Africa, and Europe, and badly in the Americas.

How places are decided

There are 6–10 places. Some of the backgrounds that fit: teachers, and people working around schools in administration, operations, curriculum, assessment, or teacher training; people with a policy background; people running education programmes or non-profits; software builders; people working on AI governance or safety; psychologists and learning researchers; and students. The list is not exhaustive. If you build things for a living and want to aim that at something more socially useful, that counts too — it is roughly how I got here.

Selection is about the mix across the group rather than ranking applications against each other.

Format

• DatesSundays 16, 23 and 30 August, and 6 September 2026
• Time12:00–13:30 IST (06:30–08:00 UTC), 90 minutes
• SizeA small cohort of 6–10 participants
• PreparationA reading and a short writing exercise before each session — roughly 60–90 minutes a week
• StyleFacilitated discussion rather than lectures
• CostParticipation is free

Possible outcomes

The group is exploratory, so there is no predetermined conclusion or deliverable.

Possible outcomes include a clearer map of the problem, new research questions, small experiments, collaborations, written reflections, or a continuing working group.

Who is behind this

I am Vatsal. One of the two questions I am actively working on is what education is for in a world transformed by AI, and how we should design it. This group is part of how I am trying to think it through, not a set of conclusions I have already reached.

I run rationality workshops, supported by a grant from the Effective Altruism Infrastructure Fund, and Portal, a community for epistemic and personal growth. I recently built an explainer on how India’s NEET exam leaked twice. I come to education as someone working it out rather than as an expert in it. More at vatsalmehra.com.