Adversarial Futures Thinking and the Endgame of AI Agents
In this exercise in adversarial futures thinking, participants will examine the capabilities of AI agents to act as both student and teacher, and design their own “teacher personas” in order to reflect on what makes their work uniquely human.
