Diary of a Language Model
What must it be like being a language model? Sharing for light relief after spending the morning with Claude refining my writing style guide, to be “more like me”.
Journal
Notes and writing
The XML feed for these posts is here.
Is Your AI Model Going Offline a Risk?
Your preferred language model going offline is not a new risk. But now that you are automating things and using agents, your approach to this risk may need rethinking.
Your Role Has Changed
There’s lots of talk about AI risks right now. Finally. But the reality in your own businesses won’t change. I recommend you embrace the opportunity, with some simple advice.
An Agent Oversight Checklist
‘Human-in-the-Loop’ checks have some serious failings. Here is a practical, non-technical checklist to use when managing AI agents.
Who Is Responsible
The NCSC published interim advice on managing the security risk of AI agents. Read it cold and you would reasonably conclude that agentic AI is only for organisations with a dedicated security team. That’s not the case; but read this first!
Skill Safety
A simplified safety checklist for anyone using AI skills
AI Transparency Is Not Enough
What do we mean by AI transparency? In the definitions emerging from policy, it is too narrow and we need to start thinking about transparency more broadly. This post is the second of three about practical AI governance, continuing from this first post.
The New Site Is Live
The placeholder website has finally been replaced and version 1 is now live. Then I did a retrospective with Claude and learnt a few things.
Designing Governance
There is a flood of AI policy masquerading as governance, and it does not survive contact with an actual AI agent. If you want to automate a process, you first have to understand that process, including the messy bits nobody wrote down. Stafford Beer worked out why this matters back in the 1970s, and his insight is uncomfortably relevant now: your controls have to be as complex as the thing they are controlling. A prompt and some context will not do it; practical governance starts with feedback loops and safe experimentation, not documents.
Vacations in Chaos
Week 31 Cartoon