I Am Often Wrong

He presents a pragmatic, six-step approach to product work - especially relevant in AI - centered on gathering and understanding information, defining the problem, choosing a clear and simple approach, setting a goal, and acting with urgency. That sequence is explicitly iterative: new data should cause repeated redefinition of the problem, approach, and goals. Churn and repeated attempts are framed as healthy parts of the process, not failures, because updating priors in light of new evidence is the only reliable path to solving complex problems.

He emphasizes a feedback-driven culture: he gives real-time feedback when people skip steps or execute poorly and expects the same in return. The most common failure modes are weak problem definition and overly complex or unclear approaches, which produce murky plans and success criteria; these faults are easier to miss without outside critique. He also signals openness to changing his own meta-process, and values being wrong because it accelerates learning, sharpens problem definition, and leads to better solutions. The note is offered as guidance to teams building product in the age of AI.