Programming instructors face a teaching crisis as generative AI tools like ChatGPT reshape how students learn to code. Three years after ChatGPT's public launch, AI remains embedded in everyday tools students already use, from web browsers to code editors. Yet research by O'Reilly Radar's collaborators reveals most computing instructors lack coherent strategies for integrating or managing AI in their classrooms.
The core problem cuts deeper than simple resistance to technology. Traditional programming pedagogy assumed students would write code from scratch, problem-solving independently. GenAI inverts this dynamic. Students now access instant solutions, explanations, and code generation with minimal effort. Instructors scramble to redesign courses on the fly, improvising what researchers call "emergency pedagogical design."
Some educators refocus on conceptual understanding rather than syntax memorization. Others emphasize debugging skills and code review. A few experiments push toward collaboration between student and AI, treating the technology as a learning partner rather than a shortcut.
The challenge remains acute because the field lacks consensus. Different institutions adopt different approaches. Some ban AI tools outright. Others embrace them fully. Most educators fall somewhere in between, uncertain whether adaptation serves students or undermines learning outcomes.
Without intentional curriculum redesign, programming instruction risks becoming either obsolete or divorced from real-world practice where AI collaboration is standard.