The Constraint Was the Brief
How treating severe constraints as the brief can clarify the real problem, expose the limits of familiar patterns, and produce a product for the user's actual state.
How treating severe constraints as the brief can clarify the real problem, expose the limits of familiar patterns, and produce a product for the user's actual state.
How AI products accumulate intent debt when they misread users, assume missing context, and make correction difficult, plus ways design teams can identify and prevent it.
How a working agent harness uses bounded delegation, explicit checks, visible failure states, and honest cost accounting to make uncertain behavior easier to inspect.
A practical framework for deciding what an early product should cut, defer, or build when ambition outruns time and budget, and research must separate evidence from enthusiasm.
How research-grounded design principles resolve real product trade-offs, survive delivery pressure, and endure when encoded in workflows, interfaces, and structure.
Lessons in research, workflow structure, delivery, and coherence from embedding with product teams across complex commercial projects and organizational change.
Why a real MVP is not the smallest demo, but the smallest complete journey a user can finish confidently, especially when trust, context, and user state shape viability.
Why designers need evals to define product quality, expose weak and inconsistent outputs, and keep AI experiences useful, clear, trustworthy, and accountable over time.
What a two-dollar agent mistake revealed about cost visibility, failure states, operational trust, and why autonomous systems must demonstrate control before earning more freedom.
Why design judgment needs systems understanding and hands-on execution, with a constrained-interface example showing how user state, constraints, and implementation shape the work.