The Variance of the Architect: Embracing the "Not Sure"

Uncertainty creates room for discovery, and also for loss.

Guarantees, estimates, and case studies are not merely demands that the future repeat the past. They can protect people from irresponsible claims and help allocate money, time, and risk. Their limits should be stated honestly when the work is genuinely new.

Experimentation becomes useful when uncertainty is defined rather than celebrated. What is unknown? What evidence could reduce that uncertainty? What is the smallest test that can answer a consequential question without exposing others to an unreasonable cost?

High-variance opportunities may offer substantial upside, but the phrase can hide the probability and distribution of harm. A failed experiment is information only if it was designed to produce interpretable evidence. It may also be a loss that deserves repair rather than a badge.

Predictable work and exploratory work need each other. Reliable operations create the capacity to test something new. Exploration can reveal practices that later become dependable.

There is no virtue in making colleagues nervous or entering a field simply because others lack a map. Leadership requires making the uncertainty visible, inviting relevant challenge, and agreeing on boundaries before the test begins.

"Not sure" is an honest starting point. It becomes strategy when joined to a question, a responsible experiment, and a willingness to stop when the evidence no longer supports the risk.