About this work
The Agentic Economy is a personal work: a thesis I have been building toward for a long time, and an attempt to render it in full. A few words on where it came from, how it was made, and what I think it represents.
The Agentic Economy is a personal work authored by Jeremy Allaire in his individual and personal capacity. Jeremy Allaire is the Chairman, Co-Founder, and CEO of Circle Internet Group, Inc., a company whose current or future products and platform may relate to topics discussed in this work. The views, opinions, and arguments expressed are solely his own. They do not represent the views of Circle Internet Group, Inc. or any of its affiliates, subsidiaries, directors, officers, employees, or investors, and should not be construed as a statement made by or on behalf of Circle. Although Circle is making the link available for convenience, the treatise is not a Circle statement, policy position, product announcement, forecast, or endorsement, and it has not been adopted or approved by Circle Internet Group, Inc., its affiliates, board of directors, or officers.
This work contains forward-looking statements, projections, and opinions about technological, economic, and regulatory developments. These reflect the author’s personal assessment and involve inherent uncertainty. Actual developments may differ materially. Nothing in this work constitutes investment, financial, legal, or regulatory advice, and no portion should be relied upon as such.
Why I built this
This treatise is the outgrowth of several decades spent building infrastructure for the internet, and of an interest I have carried since the beginning: the possibility that open software and open networks could reshape our social, political, and economic arrangements, not only the way we share information. Many of its ideas are a synthesis of thinking that first took fuller form when I founded Circle, around two convictions. The first was that money could move on open protocols, the way information already moves on the open internet. The second was that blockchains are a kind of network computer: a substrate on which autonomous software and machines could hold value, exchange it, and coordinate economic activity directly, without a person in the middle of every step.
Those origin ideas have hardened over the years into a deeper understanding of how the financial and economic system will fuse with software and the internet. And as that fusion met the arrival of genuinely capable artificial intelligence and agentic systems, it became possible to carry the thesis much further than before: to describe not merely a new kind of money or a new kind of network, but a new way the work of the economy itself comes to be organized, and the implications for humans, labor, capital, ownership and a new social contract. That is what this treatise sets out to do.
How I built this
I built it by working intensely on the core canon of my own thinking, and then collaborating with an AI to draw it out, pressure-test it, and give it form. Almost all of it began as voice (I literally have built and operated nearly 90% of this project and composition, including highly detailed technical operations, using voice controlled AI). Over the course of a few weeks I talked through the argument section by section, narrating dense blocks of thinking and then deepening each one against its strongest counterarguments. From those conversations I assembled the whole thing end to end: the written treatise at the center, and the entire experience around it.
Behind those few weeks of conversation was a great deal of machinery. The work unfolded across dozens of working sessions, and within them I directed well over a hundred specialized AI agents (drafting prose, arguing the other side of each claim, researching and checking facts, designing the diagrams, and assembling the site), often many of them working in parallel, each on a narrow piece, all converging on a single argument.
Because the subject is so large and so interconnected, I did not want it to live as a single document. I wanted a multi-modal system: a way for different people to enter the ideas at whatever depth and in whatever form suits them. You can read the full treatise or a sixty-second version of it, listen to the audiobook, watch a short film, explore the concepts as visual diagrams, or trace how they connect on a map. Every one of those forms is built from the same core written canon, and never drifts from it.
A word on the method, because it is part of the point. I used AI as an instrument, not as an author. The ideas, the judgments, the framing, and every decision about what to keep and what to cut are mine. What the tools changed was not the authorship but the reach of a single person: the ability to take an idea of this scope and carry it all the way to a finished, multi-format work, alone.
A word of honesty about the result, too. This was made by one person orchestrating many machines, not by a studio holding every component to the highest production standard, and it is not flawless. In a work this large and spread across this many forms, some pieces are more finished than others, and a system of work like this can leave fragments and artifacts behind. I have chosen to let some of that show rather than sand every edge away. The point was never a perfect object. It is that a single person could conceive, direct, and assemble the whole of it at all, and the rough edges are part of that evidence rather than a failure of it.
What it represents
What strikes me most about building this is not that one person made something large. It is how many different kinds of expert work had to come together to do it. A work like this would normally draw on an economist and a legal scholar, a research desk and a fact-checker, an editor, a brand designer and an information designer, a web engineering team, an audiobook producer, a film studio, and the people who package and distribute all of it. Each of those is a distinct discipline. In the ordinary world, each is a different profession, usually a different firm.
Here, one person directed all of them. Not by mastering each craft, but by decomposing the work into specialized cognitive capabilities and recombining them (theory and law, research and editing, design and engineering, narration and film) under a single line of intent. That is precisely the mechanism this treatise argues is coming to the economy at large: the cognitive work that once had to be bound together inside an organization, broken into parts that can be summoned, directed, and reassembled at will.
Jeremy Allaire, June 2026