How product designers can become the new front-end developer
Vibe coding tools have collapsed the gap between design and implementation. The next PM or designer can hand engineering real React code, not another handoff artefact.
Insights
Working notes on AI-native businesses.
Ideas we are writing through on product work, coding agents, enterprise architecture, security, organisational knowledge, and the operating shifts created by AI-native workflows.
Upcoming essays
From the working brainstorm.
The architecture of AI-native organisations
AI agents use the same APIs humans do. Enterprise systems now need to treat agents as first-class actors, with traceability, observability, accountability, and consent designed in.
Why Git is the right model for managing enterprise knowledge
Diffs, branches, pull requests, approvals, and history are exactly the primitives organisations need to manage living knowledge. Wikis are not enough.
Why product managers should put requirements in the codebase
AI coding agents are only as good as their context. Requirements in the repo let agents read intent alongside code and produce dramatically better work.
AI coding agents are ready for enterprise systems
The real standard for enterprise-ready is security, scale, and control. Coding agents can accelerate delivery against those standards when humans set the architecture.
If you are not using AI coding agents defensively, you are increasing your exposure
Attackers are already using agents to find vulnerabilities and move faster. Defensive agents can review code, map dependencies, and generate threat models at a scale humans cannot match.
Why product managers need to use AI coding agents to get work done
PMs who can prototype, query data, automate workflows, and draft specs with agents make better decisions faster and become more credible partners to engineering.
Rapid prototyping with AI coding agents
The time from idea to testable prototype has collapsed. That changes discovery, pitching, validation, and the threshold for deciding what is worth building.
Stopping architectural drift before it ships
Architectural drift is a governance gap. AI-readable specs and CI checks can turn the target architecture into executable policy before code reaches production.
Series arc
Building AI-native: how the work is actually changing.
Role transformation
Designers and PMs using agents to do more of the work directly.
Organisational architecture
How AI-native organisations structure knowledge, systems, context, and enforceable architecture.
Enterprise adoption
How teams adopt agents while managing the real risks around security, control, and scale.
Want to talk about one of these ideas?
Bring us the workflow, market, or operating problem you are thinking about.
Start a conversation