Episode Details

BONUS: How Scrum Masters Can Use AI Without Losing the Human in the Loop

AI makes it easier than ever to build software, prototypes, courses, and coaching tools. In this BONUS episode, Mike Lyons and Greg Pfister share what that speed changes for Scrum Masters, why product judgment becomes more important, and how coaches can start using AI without outsourcing the conversations their teams still need.

When AI Stops Being a Curiosity and Starts Saving Real Time

"It's not that the work is wrong or not needed, it is needed. That's an important step. Retrospectives are critical."

Greg's first practical AI moment came while trying to build an "Ask Mike" capability for self-paced courses. After a frustrating outsourcing attempt, he started using ChatGPT to help him rebuild the tool himself, eventually moving into Cursor, Claude Code, and the Superpowers plugin for Claude Code. Mike's moment was less technical: using AI inside Mural to affinity-map retrospective notes. The lesson for Scrum Masters is not that AI removes the work, but that it can remove enough friction to let facilitators spend more time on judgment, listening, and follow-through.

The Bottleneck Moves From Building Fast to Building the Right Thing

"The cost to produce prototypes for software engineering is approaching zero."

Mike and Greg argue that AI does not create the "wrong feature" problem, but it makes the problem much easier to multiply. If a prototype can appear before lunch, the old excuses disappear. Teams still need to ask whether the customer problem is real, whether the payoff matters, whether there is proof from users or data, and whether this work deserves priority now. For Scrum Masters and Agile coaches, this is a clear invitation to help Product Owners slow down the decision before accelerating the delivery.

Building AskMe With AI as the Engineering Partner

"I'm really playing product manager. That's really what I'm doing."

Greg describes AskMe as an AI-enabled coaching tool embedded into training courses. Instead of asking learners to pass obvious multiple-choice quizzes, AskMe asks them to apply what they learned to their own context, then reflects back practical coaching based on the course material, instructor context, and learner profile. In their own product development, Greg uses AI as an engineering partner while Mike keeps asking the product question: should we build it? Their 4P lens is simple: problem, payoff, proof, and priority.

The Scrum Master Role Becomes More Important, Not Less

"Don't just outsource your brain, your decision making power."

When leaders push teams to "adopt AI," Mike warns Scrum Masters not to let the tool become the decision maker. AI can cluster retrospective notes, summarize long threads, propose learning plans, or help prepare for a hard conversation, but the human still needs to inspect the output and understand the consequences. Greg adds the practical security angle: teams must be careful about what they paste into AI systems, especially personal, customer, or sensitive company information.

Start Small: Context, Role Play, and Shared Learning