Project-Driven AI Curriculum

Project-driven AI curriculum is the Centering humans in AI education might be key to innovation and research model described by Gaurav Sukhatme at the USC Stevens School for Computing and Artificial Intelligence. The idea is that AI teaching should be organized around projects and quickly refreshed electives so students work with changing tools and real problems rather than only fixed course sequences.

The concept extends AI University Assessment Reform and AI Default Learning Environment. If AI is already part of the learning environment, projects can make process, judgment, ethics, and domain transfer visible, while fast elective creation helps a university respond when AI topics change faster than traditional curriculum cycles.

Key Claims

• Projects help students connect AI tools to real domains instead of learning AI as isolated model vocabulary.
• Fast-refresh electives can track new AI topics more quickly than a fully fixed curriculum.
• Non-STEM minors can make AI legible to students in fields such as history and art.
• Project-based AI teaching still needs ethics, privacy, bias, and human agency built in.
• Curriculum flexibility is not enough by itself; students still need enough foundation and judgment to verify AI-assisted work.

Connections

• Gaurav Sukhatme, USC Stevens School for Computing and Artificial Intelligence, and University of Southern California - source grounding.
• Human-Centered AI Education - broader educational frame.
• AI University Assessment Reform, AI Default Learning Environment, and AI As Tutor - existing AI education branch.
• Project-Based Science Learning / 项目式科学学习 and New Engineering Education / 新工科教育 - adjacent project and engineering-education frames.
• Human Judgment Under AI, AI Shortcut Risk, and University Opportunity Density - capability, risk, and institutional-resource context.