Sue Khim, co-founder of Brilliant, is building an AI tutor designed to make students think, not just get answers. The core argument: AI cheating is a symptom of broken incentive structures in education, and the fix is not blocking the tools but redesigning how learning works.
The interview gets specific. Khim demonstrates how Brilliant's AI handles a stuck student by asking questions instead of providing solutions, a Socratic loop that forces reasoning. She identifies where US education fails concretely: overemphasis on answer retrieval, undertrained teachers, and one-size-fits-all pacing. The personalization angle is not vague. It means a student can revisit a concept at their own rate without judgment or schedule pressure.
What makes this worth watching in full is the product demo and the parent-founder tension Peter Yang brings to the conversation. He is not just interviewing Khim as a journalist. He is stress-testing her ideas as someone deciding what tools his own kids will use. The unresolved question underneath all of it: can an AI tutor scale genuine curiosity, or does it just make compliance more efficient.
[WATCH ON YOUTUBE →]