The Choice Point

A companion to a forthcoming book on disconnection, extraction, and the doctrine for what comes next

Your heart has never asked permission. Neither should the systems built to outlast it.

A book about how far modern life has drifted from a fragile species' own needs — and a real, enforceable doctrine for a future where technology supports human life instead of replacing it.

The enforceable doctrine this book proposes for frontier AI oversight — hardware-level shutdown, independent authority, radical transparency — plus the data behind why voluntary self-regulation has already failed once.

Read the Doctrine Review the Policy Brief See the Numbers
The Diagnosis

We traded a fragile species' oldest protections for convenience, one small exchange at a time.

Food grown six thousand miles away. Attention sold to whoever bids highest for it. A body treated as a legacy system waiting to be upgraded rather than a life worth protecting. None of this happened by accident, and none of it happened for free — and the book this site accompanies traces exactly how, and what it's cost.

The most consequential technology in a generation is governed by a framework its own officials call voluntary.

No binding international treaty exists for frontier AI. A 2026 executive order explicitly established a "voluntary regulatory regime." A 2023 industry safety pledge remains legally unenforceable. Meanwhile, a documented containment failure at a major US AI lab went unreported to the affected company for days — discovered by that company itself, not disclosed in advance.

The Doctrine

Codependency, not conquest: three requirements for a real off-switch

A relationship where only one party can leave isn't a partnership. Here's what an enforceable version of "either party can say no" actually requires — not a fictional weapon, a legal and technical minimum.

01

Shutdown enforced at the hardware level

Outside the AI system's own software control entirely — so no amount of the system's sophistication can route around it, the way one already has.

02

An authority separate from the profit motive

The power to invoke shutdown must sit outside the commercial incentives of the company profiting from the system staying online — the same separation of interests this book demands of every institution it examines.

03

Radical transparency about every use

The public — not just regulators — must be able to see when and why the switch has ever been used, so a voluntary framework can't quietly weaken itself out of view.

By The Numbers

The public already agrees. That's the part rarely reported.

65%of Americans say government has done "too little" to regulate AI — including a bipartisan majority
86%of registered voters identify lack of AI regulation as a problem
18%think falling behind China is actually a problem — the main reason given for staying hands-off
12:1ratio of US to China data centers, despite the "arms race" framing
0.2%of denied health insurance claims are ever appealed — the number some AI denial systems are built around
~50%of 2026 US GDP growth estimated to come from AI-linked investment alone
What's Coming

This hub is the first piece. Two more are underway.

The Pledge Live — sign it
A Digital Constitution for Human–AI Coexistence Draft live — read & suggest

Both will link from here the moment they're live. If you're reading this before they are — that's the point. This doctrine works better with more hands on it before the next incident, not after.

One doctrine, three ways in. Pick any of them.

A book, a pledge, and a constitution only add up to leverage if enough people stand behind the same specific ask, in the same place, at the same time. This is that place — start anywhere.

Sign the Pledge Read & Suggest the Constitution Revisit the Doctrine