Why I’m Still Here
I named a structural shift in 2012. AI just finished the job.
In 2012, I wrote a paper about the MicroMultinational. The argument was simple and, at the time, unpopular with anyone who had a corner office to defend: the traditional firm was being structurally dismantled. Not by a competitor. By connectivity.
The networked independent professional — someone with deep domain expertise, a laptop, and access to the same tools as a two-hundred-person agency — was beginning to deliver outsized value over bloated hierarchies that moved slower, charged more, and added process where they should have been adding judgment. I called the organism this produced the MicroMultinational. Small by headcount. Global by reach. Ungovernable by the org chart conventions that had organized professional services since the 1950s.
The Gig Economy became the cover story. But the Gig Economy was the surface. The structure underneath was the decentralization of accountability — the moment when the question “who is responsible for this decision?” stopped having a clean org-chart answer.
Fourteen years later, AI completed the job. The same structural shift that dissolved the consulting agency’s monopoly on expertise is now dissolving the firm’s monopoly on execution. An agent doesn’t clock in. It doesn’t sit in a jurisdiction. It doesn’t appear on an org chart. It makes decisions — consequential ones, at scale, continuously — and the question of who is responsible for that decision is the one every regulator in Europe, the Middle East and India is now asking in earnest.
The GDPR didn’t come from nowhere. The EU AI Act didn’t come from nowhere. They came from the structural conditions I was looking at in 2012: distributed accountability, cross-border digital services by default, and the collapse of the assumption that a human with a name and an employment contract was the unit of decision-making. When you decentralize accountability at that scale, regulatory frameworks follow. They have to. The alternative is a system with no one to sue when something goes wrong.
Structural foresight — not prophecy
The 2012 work did not “predict” GDPR or the EU AI Act. It named the structural preconditions — decentralized firms, distributed accountability, cross-border digital services — that made such frameworks inevitable. Regulators don’t write rules for centralized, single-jurisdiction firms. They write them when accountability diffuses across networks. The causal chain is structural, not prophetic.
What I didn’t anticipate in 2012 was the speed. And I didn’t anticipate the specific failure mode that would define the next wave: organizations that adopted the tools of decentralization — the APIs, the models, the agents — without building the governance layer that makes decentralized execution survivable under scrutiny. That failure mode is the subject of this book.
The Big 4 consulting agency will sell you a roadmap. A team of junior developers will spend three months figuring out an agent framework. A proof-of-concept will pass the demo and fail the regulator. The slide deck will be beautiful. The production system will not exist.
The AI Plumber doesn’t write the roadmap. The AI Plumber steps into zero, builds the governance layer — the constrained identities, the attribution chains, the kill switches that actually work — and hands you the keys to a system you can actually run.
The Fractional CAIO
The MicroMultinational’s evolved form in the AI era. Not a replacement for the CTO — the role that exists precisely because the traditional engagement model can’t move fast enough and can’t transfer ownership by design.
The tools changed. The structural problem didn’t.
This is what I’ve been doing since 2012. The chapters that follow are the plumbing manual.