The explosion of Generative AI in late 2022 has profoundly impacted society, industry, and education, demonstrating that human-made systems can master language eloquently, a feature that was thought to be unique to humans.
This shift creates a fundamental challenge to educational institutions regarding what and how to teach and to evaluate. This rises four questions: what to teach (in order for the graduate to excel in a job that probably will use AI extensively), how to teach (using AI as a tool to support instructors and students in their teaching and learning efforts), how to evaluate (in order to avoid misconduct and fraudulent practices), and what to evaluate (what skills and knowledge should be assessed). No final recipes can be given, especially when the field is evolving at exponential pace. Prompts, guardrails, skills, agents, computer use, …, new affordances emerge daily introducing an instability that makes planning very difficult. Given the current scarcity of established best practices, this presentation tries to identify some basic principles and ideas to navigate this environment, one that is both exciting and chaotic.
