The Productive Half-Life of AI Agents
Track how long AI agents stay useful before a human must step in, and design leadership rituals to extend that window without losing control.
Read on Wardley Leadership StrategiesArchive
Writing and Essays
A collection of posts, spanning engineering, leadership, and experiments.
Browse by tag
View all tags →
Blog Post
Asdfghjkl: a keyboard-first mouse controller for macOS
I built a new macOS tool in Swift called Asdfghjkl that lets you avoid the trackpad and control the mouse using the keyboard instead. It overlays a grid on the active screens, maps each cell to a letter, and keeps subdiv…
Read more
Blog Post
Cross-compiling Go for Android (Termux) With Working DNS
Go makes cross-compilation easy. Set GOOS and GOARCH, run go build, then you get a binary that Just Works. This means you can use Linux to build your project and get Windows, macOS and Android executables, across differe…
Read moreSurprises hidden in the Claude Opus 4.5 System Card
Anthropic released Claude Opus 4.5 today. You can read the official announcement, which has all the standard benchmarks, many of which it does well on. One interesting bit from the announcement caught my eye: The model’s…
Read moreThe Multi-Model Mind: Meta-Rationality for Wardley Leaders
Pairing Wardley's "Use Appropriate Methods" doctrine with meta-rationality and Munger's latticework to keep AI strategy from hardening into dogma.
Read on Wardley Leadership StrategiesAI Playbooks for Crossing the Chaos Boundary
Practical sequences to move AI components from chaotic incidents back to complex and complicated domains through minimum-viable mapping, rapid guardrails, and doctrine updates.
Read on Wardley Leadership StrategiesStrategic Entropy Budgets: Designing for Controlled Disorder in High-K Systems
Allocate explicit entropy budgets on Wardley Maps to manage NK ruggedness, focus exploration, and cap coordination cost and rework through governance levers.
Read on Wardley Leadership StrategiesExecutable Doctrine
Convert Wardley Maps and Cynefin playbooks into policy-as-code guardrails so AI agents can act within codified doctrine while humans handle exceptions.
Read on Wardley Leadership StrategiesAutonomy Gradient Maps
A new mental model that overlays Wardley Maps with autonomy bands to choreograph AI systems and human teams as capabilities evolve.
Read on Wardley Leadership Strategies
Blog Post
From Skills to Agents: Bridging Claude Skills and AGENTS.md
tl;dr: skills-to-agents automatically compiles SKILL.md into AGENTS.md. Run it as a GitHub action with a step: uses: dave1010/skills-to-agents@v1. Coding agents benefit from custom instructions and tools. The standard wa…
Read more
Blog Post
Updated site and new blog
Hello, World. This site is still static HTML but now built with 11ty and hosted on Cloudflare Pages, rather than being handwritten HTML on GitHub Pages. This means it's easier for me to add content and will hopefully res…
Read more
Elsewhere
Interactive Planning, Idealised Design, and Wardley Mapping
Use Ackoff's interactive planning to pair idealised design with Wardley Maps, turning a shared vision of the future into executable gameplay.
Read on Wardley Leadership StrategiesSoft Systems Methodology Meets Wardley Mapping
Use Checkland's Soft Systems Methodology to clarify messy problem spaces before mapping value chains with intent.
Read on Wardley Leadership StrategiesDouble-Loop Learning Keeps Wardley Maps Honest
Wardley Maps feel definitive, but double-loop learning stops them from ossifying into doctrine and keeps teams adaptive.
Read on Wardley Leadership Strategies
Elsewhere
The Cybernetic Fate of Organisations
Wardley Mapping the conserved risk profiles organisations face when they attenuate or amplify variety.
Read on Wardley Leadership StrategiesPanarchy, Adaptive Cycles, and Wardley Climatic Patterns
Use Panarchy to read the resilience dynamics hiding behind Wardley Map evolution, turning climatic inevitabilities into strategic timing signals.
Read on Wardley Leadership StrategiesRugged Landscapes and Wardley Maps
Using Kauffman's NK model to understand why modularity, doctrine, and exploratory gameplay matter when mapping complex systems.
Read on Wardley Leadership StrategiesCybernetic AI Leadership with the Viable System Model
What happens when leaders wire Stafford Beer's Viable System Model into their AI operating system?
Read on Wardley Leadership StrategiesAI-Accelerated User Needs Leadership
Treat large language models as co-analysts who keep Wardley Maps anchored on layered user needs instead of just today’s feature backlog.
Read on Wardley Leadership StrategiesLLM-Driven Competitor Simulations
Using language models to stress-test your Wardley Maps against plausible competitor plays without drowning in hypotheticals.
Read on Wardley Leadership StrategiesBackground AI for Relentless Improvement
How ambient agents refactor architectures, uplift process maturity, and accelerate Wardley transitions without demanding heroics.
Read on Wardley Leadership Strategies
Elsewhere
Positioning Readiness for the AI Age
Why leaders must master map positioning and build multi-play readiness as AI agents collapse the cost of optionality.
Read on Wardley Leadership Strategies
Elsewhere
Winning AI Leadership Cycles with the OODA Loop
How can AI-era leaders wield the full OODA loop to outpace autonomous competitors without losing ethical control?
Read on Wardley Leadership Strategies
Elsewhere
The Collapse of Differentiation in the AI Red Ocean
Why Wardley Mapping leaders must anticipate the accelerating evolution of software components.
Read on Wardley Leadership Strategies
Elsewhere
The Age of Diffused Agency
How AI-driven agency is shifting execution power to individuals and redefining leadership priorities.
Read on Wardley Leadership Strategies