AI Education

EdTech AI promised to revolutionize learning; reality is messier. Chegg's Snapchat-for-homework model faces disruption from Claude and ChatGPT—students now go straight to LLMs for answers. Traditional platforms like Coursera and Udemy have added AI tutors, while startups like Duolingo (AI-driven personalization) command premium engagement metrics. Khan Academy's Khanmigo targets K-12, using AI to play Socratic guide. The fundamental tension: AI tutors are cheap and available 24/7, but they lack the motivational power and accountability of human educators. The winning startups pair AI tutoring with social/competitive features (Duolingo's streaks and leaderboards), or focus on high-value niches where expertise is scarce and expensive (pilot training, medical board prep, coding bootcamps). Regulatory questions are mounting: state education boards are grappling with AI-generated homework, and parents worry about over-reliance on AI. The defensible winners will be those embedding AI as a tutor alongside human instruction, not replacing it. Personalized learning is AI's killer app in education, but only if paired with outcomes data and human accountability.