Frame the real outcome
Name the person, decision, artefact, or behaviour that should change when the work succeeds.
Available now
The field guide for modern AI work
The AI Operator is a practical system for thinking, prompting, researching, creating, coding, automating, and building with ChatGPT, Codex, and specialised AI systems such as KAI.
30 chapters, 192 reusable prompts, 9 diagrams, and practical examples across work and life.
From access to capability
A capable operator does not depend on one clever prompt. The operator frames the outcome, supplies source truth, chooses the right working surface, checks the evidence, and preserves what works.
Name the person, decision, artefact, or behaviour that should change when the work succeeds.
Bring forward the facts, examples, files, constraints, and permissions that should govern the result.
Review the actual file, calculation, code, page, or customer journey before treating the work as complete.
The Operator Loop
Use the same loop for a short email, a source-backed report, a Codex project, or a customer-facing system such as KAI.
Learn the complete methodDefine the outcome and audience.
Supply source truth and constraints.
Produce a visible first result.
Check evidence and acceptance tests.
Correct the work with specific feedback.
Save the approved process for reuse.
Route first, prompt second
Use dialogue for explanation, exploration, brainstorming, rehearsal, and iterative thinking.
Think and createUse project access for code, files, commands, diffs, tests, and verified implementation.
Build and verifyUse a specialised system for a bounded audience, approved knowledge, guardrails, and human handoff.
Guide and hand offInside the book
The book moves from mental models and prompting into tools, business systems, safety, scale, and future capability.
Explore all 30 chapters5 chapters
6 chapters
9 chapters
5 chapters
5 chapters
A practical prompt library
The prompts cover business, marketing, research, writing, coding, education, automation, productivity, sales, customer support, image generation, data analysis, and specialised AI systems.
Preview the libraryBBackground What the assistant needs to know.
RResult The change the work must produce.
IInputs The facts, files, examples, and sources.
EExecution The process, constraints, and permissions.
FFinal checks The tests that define acceptable work.
The KAI pattern
KAI appears throughout the book as a practical example of an assistant built around a narrow job, approved knowledge, honest boundaries, and a clear human handoff.
Questions
It is written for founders, professionals, marketers, creators, educators, developers, and anyone who wants AI to produce useful, dependable work rather than more unverified output.
No. The book starts with practical mental models and everyday ChatGPT work, then introduces more capable workflows such as Codex, automation, and specialised systems at a measured pace.
The 192 prompts sit inside an operating method. You learn how to frame outcomes, choose the right tool, ground work in evidence, inspect results, improve them, and preserve the approved process.
KAI appears throughout as a practical example of a specialised AI system with an approved knowledge base, a bounded customer-facing job, clear guardrails, and human handoff.
Kindle, paperback, and hardcover editions are being prepared through Amazon KDP. The Kindle edition is intended for Kindle Unlimited through KDP Select.