Kahneman's Humility and Open-Mindedness
by Annie Duke on May 2, 2024
Annie Duke believes making implicit decision processes explicit is essential for improving decision quality. Her approach centers on structured independence before collaboration, challenging the common belief that alignment should be the goal of decision-making.
The most impactful change organizations can make is separating the discovery of ideas from their discussion. Duke explains that meetings should only be for discussion, not for generating ideas or making decisions. When people brainstorm or share opinions in real-time, the loudest or most senior voices dominate, creating a coercive environment that narrows the range of possibilities.
Instead, she advocates for "nominal groups" where people work independently first. Before meetings, have team members submit their thoughts, forecasts, or evaluations separately without seeing others' responses. This approach reveals the true diversity of perspectives and prevents conformity bias. When you gather afterward to discuss, focus on areas of disagreement rather than wasting time restating agreements.
This independence-first approach works for various decisions - from product roadmaps to timelines, budgets, and strategic choices. The process can be as simple as having everyone write down their estimates before sharing, or as structured as using forms and rubrics for recurring decisions.
Duke emphasizes that alignment is an unrealistic and counterproductive goal. Real teams have diverse perspectives that won't perfectly align, and pretending otherwise creates pressure to conform rather than contribute honestly. Instead, leaders should make people feel heard by reflecting back their perspectives without judgment, then make decisions with the understanding that not everyone will agree - and that's perfectly fine.
For long-term decisions, Duke challenges the notion of "long feedback loops." She argues there's no such thing as a truly long feedback loop - only our choice to wait for final outcomes rather than tracking leading indicators. By identifying necessary intermediate signals that correlate with desired outcomes, teams can get feedback much sooner and adjust course accordingly.
When implementing premortems, Duke stresses they're only valuable when paired with pre-commitments. Rather than just imagining what could go wrong, teams should establish specific kill criteria - signals that would trigger pivoting or stopping - and commit to actions they'll take when those signals appear. This helps overcome our natural reluctance to quit failing initiatives.
The underlying principle across all these approaches is making the implicit explicit. Our intuition isn't always wrong, but it isn't always right either. By surfacing assumptions, creating structure, and committing to actions in advance, we can discover which intuitions serve us well and which lead us astray.