Reflections
Observations from practice.
Short essays on design, organisations, and intelligent systems.
Drawn from years of designing enterprise software, public services, and intelligent systems, they explore recurring themes of trust, understanding, complexity, accountability, and change.
Each piece approaches a different question.
Together, they explore the same one:
How do people remain capable of acting in systems they can no longer fully understand?
2026-09 · 4 min read
Writing rules for a game nobody knows they're playing
AI agents can read components and rules. The harder part is making the judgment behind those rules available when they generate.
2026-09 · 4 min read
The work that disappears when it works
DesignOps succeeds when friction disappears. But if the work itself becomes unobservable, organizations lose the evidence they need to sustain it.
2026-08 · 2 min read
The work moves upstream
Generating alternatives is becoming remarkably inexpensive. Deciding between them is not. That changes where design happens.
2026-07 · 3 min read
The designer disappears
As execution becomes automated, the visible outputs of design become easier to generate. What remains difficult is everything that happens before those outputs exist.
2026-07 · 3 min read
The design of beginning
Impressive AI demos create attention. Good beginnings do something harder: they make a new capability feel close enough to try.
2026-06 · 3 min read
When interaction becomes governance
Agentic systems don't just execute instructions. They adapt. That changes what feedback is for — and who is responsible for what the system learns.
2026-06 · 3 min read
We know more than we can explain
For a long time, expertise meant understanding. Intelligent systems are quietly separating those two things.
2026-05 · 4 min read
Capability scales. Understanding doesn't.
Most AI demos are evaluated by whether the system completed the task. That is rarely the most important question.
2026-04 · 4 min read
Trust through constraints
Intelligent systems become trustworthy through their limits, not their capabilities. A note on why boundaries are the load-bearing part.
2026-04 · 3 min read
Trust is a design material
Agentic systems don't just need good interfaces. They need to be believable. That's a different design problem entirely.