New to AI?
A General Primer for Operators Considering AI
A practical introduction to what AI is, how generative AI fits into the broader picture, where current systems help, where they fail, and why the surrounding interaction matters.
AI Interaction Methodology
A structured approach to working with AI on serious tasks, focused on runtime stability, artifact separation, and disciplined interaction for solo expert operators.
Where to start
Considering AI for real work?
A Practical Primer for Operators Considering AI
A grounded look at where AI is useful in actual work, why casual use often disappoints, and how disciplined interaction changes the result.
Ready to use the methodology?
An Operator Guide to the AI Interaction Methodology
A procedural guide to using the methodology in practice: how to bootstrap the canonical core, structure work, manage uncertainty, and maintain control.
Want the canonical source?
Explore the Public Core Repository
The public canonical source for the AI Interaction Methodology, including the methodology, framework, collaboration guidelines, and supporting artifacts.
What this is
The AI Interaction Methodology is a structured approach to working with AI on serious tasks.
It did not begin as an attempt to create a methodology. It began as a practical attempt to make AI interaction more usable for real work. Over time, it became clear that the deeper problem was not just prompt quality. It was how to make the interaction behave more like a controlled system and less like an improvisation.
This is an attempt to address that problem.
Why it exists
A great deal of AI use still assumes that better outcomes come mainly from better prompts.
That can help at the margins. But once the work becomes more serious, the failure modes change. The problem is no longer just whether the system can produce something plausible. The problem becomes whether the interaction can remain coherent under ambiguity, iteration, conflicting inputs, and shifting goals.
At that point, the questions become:
- What is authoritative?
- How is the interaction assembled?
- What defines the reasoning structure?
- What belongs to collaboration behavior rather than reasoning?
- How do you extend the system without quietly redefining it?
The methodology exists to address that layer of the problem.
What it consists of
The public core is organized around three distinct artifacts:
- Methodology — the runtime contract
- Framework — the reasoning structure
- Guidelines — the collaboration defaults
These three artifacts are related, but they are not interchangeable. Keeping them separate is part of the design.
What comes after this
The public core is the stable foundation, not the full body of work.
The pages linked above are meant to help readers enter at the right level: broad orientation first, practical grounding second, operational use third, and canonical source last.