Rob Collie releases Fair Game on custom AI for business

Rob Collie, founder and CEO of P3 Adaptive and a founding engineer behind Power BI, released Fair Game: Customizing AI to Your Business Is Easier Than You Think today from Holy Macro! Books. The book argues that companies get better AI results by focusing on data, instructions and ordinary software — and by putting the work in the hands of internal “Crafters.”

Why it matters: - Fair Game targets a gap in the AI market: business leaders want results, but many are still buying generic AI tools without the data and workflow support needed to make them useful. - The book argues that custom AI is less technical than many executives assume, which could lower the barrier for companies trying to adopt AI internally. - The release also doubles as a case study. P3 Adaptive remade its own business around custom AI and delivered that approach to clients.

What happened: - Rob Collie, CEO of P3 Adaptive and a founding engineer behind Power BI, released Fair Game: Customizing AI to Your Business Is Easier Than You Think today through Holy Macro! Books. - The book is available now at FairGameBook.ai and will also be sold through Amazon, Barnes & Noble, Books-A-Million and Porchlight Book Company. - Formats include hardcover, audio and e-book. - Collie also used a custom AI editor, nicknamed Eddie and built on Anthropic’s Claude, while writing the book.

The details: - Collie wrote Fair Game after deciding in late 2024 to shift his company toward custom AI. - He says he could not find the basic explainers he needed, so he built them for himself and his team. - The book’s core claim is that AI success depends less on the model itself than on the surrounding data, instructions and software that teach the model how a specific business works. - Industry analysts have recently labeled that work “context engineering.” - Fair Game is written for business readers rather than engineers and explains why off-the-shelf AI often falls short inside companies. - Collie argues that the people best suited to lead this work are internal “Crafters,” which he describes as roughly one in sixteen knowledge workers who build spreadsheets, automate processes and solve problems IT does not reach. - Collie says the real opportunity is teaching AI how a business actually works and empowering the right people inside the company to do it. - Eddie challenged drafts, held Collie to his arguments and caught inconsistencies across chapters. - Collie says Eddie required no LLM development or research, only data, instructions and ordinary software. - Early endorsements came from Diego Oppenheimer, John Hancock, Mike Nichols and Jamie Cool. - Oppenheimer called off-the-shelf AI “a demo” and custom AI “the business.” - Hancock called the book “the roadmap.” - Nichols said it “should be required reading in every MBA program.” - Cool called it “the clearest explanation for non-engineers.” - Advance copies are already circulating inside Fortune 500 leadership teams, and some early readers have ordered extra copies for executives.

Between the lines: - Collie is positioning the book as a practical counterweight to hype-heavy AI commentary from journalists, academics and strategy consultants. - The book’s message suggests AI adoption may be less about breakthrough technology and more about operational discipline, internal talent and business context. - The Eddie example is meant to reinforce the book’s thesis by showing that custom AI can be built with familiar tools and clear instructions, not just specialized research teams.

What's next: - Fair Game is expected to reach business leaders through bookstore sales, direct orders and executive circulation of advance copies. - Collie’s broader pitch is that more companies can build useful custom AI systems if they identify the right internal operators and give them the right data and software support. - The book may also extend P3 Adaptive’s visibility as a firm focused on data and AI implementation for mid-market and Fortune 1000 clients.

The bottom line: - Fair Game argues that AI becomes valuable when businesses stop treating it like a standalone product and start treating it like a company-specific system.