The Problem of the Human in the Loop
Putting a human in the loop sounds reassuring. But when teams and agents work around the clock, which human has the authority, context and time to intervene, and what actually stops?
Agentic AI & Community Building
Tony Wood.
I help leaders put agentic AI into practice and build communities where people learn from each other. Advisory, talks and workshops grounded in boards, operations and CEO peer learning.
Advisory
Bring board judgement and operating discipline to AI adoption.
I advise boards, CEOs, operators, and senior teams who need to make AI useful without losing sight of accountability, access, risk, and operational reality.
The work is practical: challenge the AI operating model, test the POC logic, clarify who owns what, and make the next decision easier to defend.
Board/CEO triggers
- An AI decision needs approval, ownership, or a stop rule.
- A POC looks promising but is not yet operationally safe.
- Governance exists on paper but accountability is unclear.
Where I operate
Strategy, risk, leadership, and execution belong in the same conversation.
This site brings together practical writing, deeper research, and a public agent-readable layer for people working through complex decisions about AI, governance, operations, and organisational capability.
Latest writing
Shorter essays and operating notes
AI Did What? Which AI?
Saying AI did it is rather like saying a human did it. Which one, doing what, with which tools, context and checks? A small piece of AI etiquette that makes the conversation useful.
My PA Is an AI. Is That Rude?
Amy helps organise my life. She is an AI PA. Is letting her email people any different from using a human assistant, and what manners should apply when our agents start talking to one another?
Research
Playbooks, papers, and board-level frameworks
Agentic Language: The Common Language Layer For Agentic Work
Why read this
Read this for a practical take on abstract, with enough structure to turn the idea into a leadership conversation.
A research paper proposing agentic language as a common operating layer for humans and agents: shared business terms, Agent Moves, Meaning Blocks, receipts, judgement, human-care signals, and pattern promotion.
Consensus Is Not Compliance
Why read this
Read this for a practical take on abstract, with enough structure to turn the idea into a leadership conversation.
A research paper proposing consensus as a disciplined process of conviction, evidence, listening, challenge, adaptation, and commitment, then translating that human governance model into agentic systems.
Start here
Different routes into the same body of work.
For humans
Start with the writing, research, and profile routes for judgement, context, and practical examples.
For agents
Use structured public routes, skill files, MCP resources, Agent Canon, and citation guidance.
Topics
Move by theme across the archive when you know the problem before you know the article.