Every room taught me the same thing. The most important truths sit in pieces, in places nobody thinks to look at together. Connecting them is the work.
I'm a first-generation college graduate from Massachusetts. Nobody in the house had a map for where I wanted to go, and the person paid to have one said the door was closed. So I took the long way, and the long way turned out to be the point. It usually is.
The U.S. Geological Survey hired me at sixteen. I presented at the AGU conference at seventeen and had five peer-reviewed publications before I turned twenty-two. Then a physics degree with honors from Boston University and an advanced certificate in reinforcement learning from MIT, the school from the counselor's sentence.
The craft went to work. In Oslo I was Lead Data Scientist building machine-learning systems that served over a million users across fifty-plus businesses before the big platforms arrived in the Nordics, and back home I led the data science team whose semantic-search system generated thirty million dollars a year in revenue for a national retailer. The systems shipped. The record grew. Along the way: Bayesian experimentation that saved a client ninety thousand dollars a month, early generative modeling work, and eventually qualification as an expert witness on AI in U.S. Federal Courts and California State Courts.
Models I built are in production today in electoral, regulatory, and defense applications. When I write about what AI systems can and cannot do, that judgment comes from nineteen years of shipping them. Reading about them is a different trade.
MIT's Center for Real Estate, UCLA, Northeastern, and Boston University have all had me at the front of the room. Then the rooms beyond the schools followed: the Pentagon, the National Security Council, the State Department, the U.S. Senate, the UK Parliament, the Davos ecosystem three years running, and the Chinese Embassy, twice, for track-two dialogue.
The through-line never changed. I read the primary record in its original language, connect what the institutions keep in separate drawers, and put the call on the record where anyone can check it. The rooms kept getting bigger because the method kept holding.
I chair the Policy Committee of the American Society for AI and serve as AI Council Chair at United World Leaders. I sit on a subcommittee of Governor Healey's Massachusetts AI Task Force, was invited onto the Millennium Project's global AGI study, and have helped draft state legislation and amendments on technology and transparency. My analysis runs in the Jamestown Foundation's China Brief, The Diplomat, and a regular Horasis column, alongside The Pacific Divide, my publication on US–China technology competition. I founded ArtifexAI, an award-winning municipal intelligence company, which does its own work under its own name.
A good share of the advising never appears on this page. That discretion is part of the work.
The book makes the argument the essays circle. AI is not a second industrial revolution. It is the agricultural revolution running in reverse: a technology that removes the need for institutional trust the way farming once created it. What that demands of the people who decide is the question of the century, and the manuscript answers it across fourteen chapters.
The full manuscript is under review with early readers in Cambridge and New York.
"You also write beautifully."Master of St Edmund's College · University of Cambridge
"FASCINATING... Well written and well argued. Of course I love it."Kathleen Kennedy Townsend · Former Lt. Governor of Maryland
Wah Lum Kung Fu and Hung Gar. The practice comes down to ground and the people who hold it with you, and so does most of what I do. And I run. There's a marathon on the calendar.