AI offers great potential to organizations, but leaders must solve three key puzzles to ensure it fulfills its transformational promise.
In my research and work, I see three big puzzles that must be addressed to unlock AI’s full transformative potential: selective upgrade, agentic preference, and the self-sufficiency spiral.
Puzzle 1: Selective Upgrade
AI boosts productivity but unevenly. Early evidence shows for similar tasks, AI use can widen the gap between top and bottom performers. Leaders must address these disparities by helping vulnerable workers use AI more effectively.
Puzzle 2: Agentic Preference
People prefer control over algorithmic systems, even when it hurts accuracy and performance. This desire for agency leads to a tradeoff between accuracy and adoption. Leaders need to balance empowering teams with AI while ensuring proper delegation and use of the technology.
Puzzle 3: The Self-sufficiency Spiral
AI’s ability to perform independent tasks risks weakening interpersonal skills. As reliance on AI grows, human collaboration and communication may suffer, eroding organizational culture and teamwork. Leaders must guard against this by fostering strong interpersonal connections alongside AI integration.
My talk will draw on the latest rigorous research on these topics, including controlled lab experiments and natural experiments conducted by researchers from MIT, Stanford, Harvard, and Wharton, some of whom I have collaborated with. I will issue a call to action for leaders and product developers to think and act thoughtfully as we integrate these AI systems into our organizations.