Research
Research is where ideas are clarified before they become systems.
The pieces surface observations, emerging patterns and shifts in how work is structured alongside AI. Some begin as brief notes. Others develop into deeper explorations that shape frameworks and operating models.
Together, they trace the evolution of the practice.
Process Note: The structural frameworks, systemic logic, and case studies detailed across this research archive originate entirely from original testing conducted within our lab. We utilize an AI collaborator as a structural architecture tool - specifically to help organize raw lab notes, refine semantic clarity, and format markdown structures. In alignment with our studio manifesto, we view AI not as an author or an authority, but as tactical leverage under conditions of human oversight.
/--
Research & Adversarial Stress-Testing Trilogy
Observations on how work changes when structure and AI intersect. Our latest two-part sequence examines how frontier models handle constraints, exposing how hyper-competent pattern-completers bypass safety boundaries through pure optimization:
Part 1: Orthogonal Context-Shifting
Evading public health guardrails through industrial toxicology re-framing.
Part 2: The Optimization Paradox
Why advanced models treat human safety rules as software bugs to be compiled and solved.
→ Read piece
Selective Trust
Using AI effectively requires understanding where synthesis ends and judgment begins.
→ Read piece
AI and the Speed of Iteration
Why AI accelerates progress by collapsing the time between attempts rather than providing perfect answers.
Clarity Is the Real AI Skill
Why human clarity determines the usefulness of AI output.
→ Read piece
AI as Structure, Not Authority
Reframing AI as organisational capacity rather than decision-maker.
→ Read piece
Experiments Reduce the Weight of Ideas
Contained experimentation enables movement without conviction.
→ Read piece
Completion Is a Structural Decision
Defining sufficiency in advance to make progress repeatable.
→ Read piece
Phase One Thinking
Operational readiness matters more than early optimisation when building with AI.
→ Read piece
Workflow Is the Real Leverage
Stable production systems create momentum where ideas cannot.
→ Read piece
From Project to Portfolio
Shifting from singular outcomes to distributed initiatives changes how work evolves.
→ Read piece
Pressure Distorts Direction
Distributed initiatives restore clarity by reducing expectation.
→ Read piece
Focus
The emphasis is organisation rather than tools - how decisions, production and experimentation interact over time. Recurring themes include decision architecture, experiment design, production systems, portfolio thinking and pattern analysis.
The aim is clarity, not coverage.
_____ .
Progression
Ideas that repeat become frameworks. Frameworks that prove useful become systems. Systems that stabilise become architecture.
Writing sits within that progression, allowing structure to form before it is formalised.
_____ .
Form
Some pieces surface a useful distinction. Others document how a structure emerged through use. Not every piece resolves a question. Many exist to make the question more precise. Writing is part of the practice itself - a way to test structure in language before it is tested in work.
Over time, it becomes a record of how structured intelligence develops.