Why AI agents need SOPs

Illustration of a documentation specialist reviewing an SOP checklist beside an AI agent workflow with approval steps.

AI agents are starting to do real work. They can search systems, draft answers, update records, test software, summarise documents, create pull requests, and trigger other tools. In some organisations, they are already moving from experiment to everyday workflow. That causes a documentation challenge. If an AI agent is going to act on behalf of a… Read more »

Claude Tag and technical writing: useful assistant or knowledge trap?

Last week, Anthropic launched Claude Tag for Slack. For documentation teams, it is worth paying attention to. At first glance, it looks like another AI assistant added to a workplace tool. You mention @Claude in a Slack channel, ask it to do something, and it replies in the thread. That undersells the change. Claude Tag… Read more »

Open Knowledge Format: What it means for technical documentation

graphic titled “Open Knowledge Format (OKF): What it means for technical documentation”, showing three connected Markdown-style document cards with YAML metadata, a central knowledge graph icon, and a subtle AI network illustration.

Could technical documentation become the knowledge layer for AI? This month, Google Cloud has introduced the Open Knowledge Format (OKF). It’s an open specification for packaging organisational knowledge in a form that both people and AI agents can use. An OKF collection is essentially a directory of Markdown files. Each file contains a small amount… Read more »

Five years of getting AI wrong

Illustration titled “Five years of getting AI wrong.” A winding, ascending stone pathway shows stages of AI use: 1) Prompt engineering—craft the input, get the output; 3) Vibe coding—describe an application and receive working code; 4) Agents and orchestration—chain tools and automate workflows; and 5) Goal-directed AI—describe the desired outcome and let AI determine the steps. People work at laptops along the path, while a woman in the foreground looks toward a bright summit labeled, “What are we actually trying to do?”

Our first experience of AI began with documenting AI systems in the early 2020s. One product guided banking customers using spoken interactions through a decision tree, and another vectorised data. We thought we understood what AI did, and where it could be used. Then, in 2022, OpenAI released ChatGPT. In five days, ChatGPT had a… Read more »