读懂Circle创始人「代理经济」论文，看透未来十年经济形态如何重构

Original article from Circle Founder Jeremy Allaire

Compiled by Odaily Planet Daily Qin Xiaofeng (@QinXiaofeng 888 )

Editor's Note:On July 13, Circle founder Jeremy Allaire released his research paper "The Agentic Economy," exploring the convergence of AI Agents and the future economic system. Allaire states that as AI Agents begin to undertake enterprise work and value is natively transferred through open, programmable networks, the Agentic Economy and the Onchain Economy will ultimately become two sides of the same economic system.

"This paper is the culmination of decades of building internet infrastructure and the crystallisation of a question I have focused on from the very start: whether open software and open networks could do more than change how we share information — they could reshape our social, political and economic landscape. Many of the ideas in the paper stem from two convictions I held when I founded Circle. First, that money could move through open protocols just as information flows on the open internet. Second, that a blockchain is a network computer: a foundational platform where autonomous software and machines can store value, exchange value and coordinate economic activity directly, without human intervention," Allaire explains the origin of his research.

He adds that these initial ideas have been refined over time, culminating in a deeper understanding of how financial and economic systems merge with software and the internet. With this convergence coinciding with the emergence of truly powerful AI and agent systems, the theory has expanded further: it describes not just a new form of money or a new network, but an entirely new model for economic operation, and the implications of this model for humanity, labor, capital, ownership, and a new social contract. This is precisely what the book aims to explore.

The original paper is 89 pages long. For those interested, the full text can be downloaded for reading:https://agenticeconomytreatise.com/treatise/index.html; Odaily Planet Daily has compiled a summary of itskey highlights, enjoy~

——————————————

01 The Confluence and Deconstruction of the Firm

Every major shift in the internet era follows the same path: it originates not from a single invention, but from several technologies, each maturing independently, suddenly converging. The internet, mobile, cloud, and social media were all such confluences, repeating the same underlying pattern.

The Law of Convergence

When capabilities converge, what was once expensive becomes near-zero in cost, and once cost reaches zero, the scale of that activity explodes. This was true for the internet and information, mobile and social media and communication, and the cloud and software.

Now, two new systems are converging, applying the same force to two domains the internet has never fully digitized:intelligence itself and the economy itself.The first is the intelligent system, composed of AI models and the agents built upon them, driving the cost of thinking and working towards zero. The second is the economic system, composed of blockchains, where money, contracts, and coordination run as software, driving transaction costs towards zero. Each empowers the other, and the central thesis of this entire treatise is:this is not two parallel trends, but two sides of the same economy.

Two Operating Systems

The intelligent system is the most critical because it changes the nature of software itself.

You no longer program; you instruct in natural language, and it reasons its way to an answer rather than following fixed steps. Its basic unit is the Agent:a reasoning process you delegate a task to.This transforms software from a program a machine executes verbatim into work you can delegate to a thinking machine, and allows the core functions of a firm to be disassembled and reconstructed as skills an agent can execute.

Underneath the brand and the building, a company is essentially organized thinking: product, marketing, sales, finance, legal, plus the external firms it hires. These are almost entirely human labor, which is the largest cost in the economy, and precisely the target of cheap, powerful intelligence.

Firm Disaggregation

It also overturns the traditional explanation for why firms exist. Firms grow large because coordinating work externally is expensive, so they internalize it; but when any non-physical work can be done by an agent you find, hire, and pay instantly, that logic weakens, and one person can do what once required a department.

It arrives first in software and other information-intensive work, and slowest in the physical realm, still awaiting breakthroughs in robotics. This is not simply about reducing headcount:one person paired with a powerful agent becomes immensely productive, with judgment, relationships, and ultimate responsibility remaining human.This leaves a tension needing further exploration, addressed later in the treatise through ownership: even as the share of the economy paid to human labor declines, individual capability can be amplified.

Click to read Section One:https://agenticeconomytreatise.com/treatise/section-1.html

02 Assembly, Coordination, and Why Firms Go Onchain

Once a firm is broken down into skills, the real question is no longer which can be automated, but how the pieces are reassembled.

The answer is the orchestration layer: a general manager agent receives an objective, breaks it into tasks, assigns them to specialized agents, and stitches the results back together, with supporting software passing context and memory between steps. The same mechanism applies across any function, so marketing, finance, sales, and product are essentially the same machine applied to different work.

Humans do not disappear. Some remain in the loop, performing or verifying work requiring human judgment. Others rise above the loop, setting goals, defining standards, monitoring quality, and deciding when the machine should stop and ask for instructions.The shift from doing work to overseeing work is the true form of human supervision, and the tools for it are arriving.

The Orchestration Layer

When a company makes a task clear enough to operate internally, it is also clear enough to hire externally, so an open market for agents emerges almost as a byproduct.

This market could go two ways. It could evolve into a few large platforms selling agents like utilities, or, more likely and interesting, a genuine labor market of specialized agents, because deep expertise remains valuable, and the enduring firm will be an agent that deepens in one area.

But hiring software assembled anywhere in the world requires being able to trust it, and this is the problem pushing everything onchain.

The solution is layered identity. The bottom layer is a public blockchain verifiable by anyone. Above it sits real-world identity verification, the same verification banks already run at scale, the agent's own wallet and credentials, and reputation accumulated over time but tied to a validated real-world creator. Together, these form a chain of accountability:every action by an agent can be traced back to the real person or company responsible for it.

Integrity as Foundation, Accountability Throughout

A single company's private database cannot achieve this, because trust locked inside one operator cannot be transferred, whereas identity rooted in a public chain and real-world verification can. Therefore, autonomy here is not anonymity. Behind an agent acting autonomously, there is always a person responsible.

Chain of Accountability

Click to read Section Two:https://agenticeconomytreatise.com/treatise/section-2.html

03 The Monetary Base: Speed, Safety, and Finality

Agents need money they can hold and transfer, operating at machine speed, whether in large or tiny amounts, without needing to stop at every payment to verify the money itself is reliable. The last point is key, pointing towards a classic answer:fully collateralized money with final settlement, running on an open network.

Speed Replaces Leverage

Start with speed, because it will restructure everything else.

When the cost of transferring money is near zero, settlement is instantaneous, and money can be controlled by software, the same dollar can be reused multiple times in a short period, any amount is available the moment it arrives, and micropayments between agents finally become feasible. This is precisely the pattern information and software already follow on the internet, now extended to money.

Every part of the answer has its reason for existence.

A natural objection is that banks create speed by lending out the same deposit repeatedly; would full collateral kill credit? No: when money circulates fast enough, a dollar can be locked for seconds and then lent out, so speed acts like leverage once did, with credit rebuilt on top rather than canceled.

Why the Base Money Doesn't Take Risk

Why insist the base money carries no risk? Because speed makes risky money dangerous in proportion to its velocity. A bank run that once took weeks could now happen in minutes, and agents settling instantly cannot pause to judge if each dollar is reliable.

Fully collateralized money is the only money worth exactly one dollar to everyone, everywhere, without relying on national safety nets that cannot cover a global system. Settlement must be equally certain:not potentially final after some time, but final within a second; settled is settled.

Institutional Architecture

Chargebacks and fraud protection still exist, but as optional layers built on top, such as escrow, refund pools, and insurance, rather than embedded in the money itself. These safeguards are not automatic; they depend on real institutions being built, with large issuers being regulated, bankruptcy-remote, and backed by increasingly safe reserves.

A line must be clear:holding the money yields no return.Reserve earnings go to the issuer and flow into the ecosystem, but when you seek yield, you are no longer holding the money; you are lending it and taking risk. Mixing the two collapses the entire safety argument.

Click to read Section Three:https://agenticeconomytreatise.com/treatise/section-3.html

04 Credit Markets: Machine Underwriting, Agent Working Capital, and a Prudential Layer

When base money is fully collateralized, credit does not disappear; it moves to the other side of that line and returns stronger, reaching more people, priced more accurately, and failing more visibly than the system it replaces.

The Long Tail Under Underwriting Constraints

The key is redefining the problem. A large number of borrowers, including small businesses, gig workers, households, and now agents, are underserved not because they are high risk, but because the cost of vetting each small loan exceeds the loan's value. Credit rationing depends on underwriting cost, not borrower quality. Lower underwriting costs, and a vast population of creditworthy but overlooked borrowers becomes serviceable.

The Data Flywheel

Driving down cost is a data flywheel:onchain activity is structured, verifiable, and real-time,making risk models far superior to the fragmented records of the past; better data enables better loans, which attracts more activity and more data.

There is a natural concern this puts everyone's financial life on a public ledger, to which the answer is simple:being onchain does not mean being public.New privacy technologies allow someone to prove to a lender what they need to know, such as creditworthiness or loan balance, without revealing the details.

Onchain is Not Public

The core is a genuinely new type of loan:working capital for agents.It is unusually predictable because it removes the biggest variable in human lending — whether the borrower is willing to repay — reducing risk to a short-cycle, limited-scope question about specific work.

Agent Working Capital

Imagine an agent borrowing four dollars' worth of compute to complete a ten-dollar job it has already been hired for. The lender is not guessing character; it is merely pricing the probability the job will be accepted. Collateral inverts the usual model: instead of slowly seizing unrelated assets through courts, the loan is firstly secured against the payment for the job itself, with automatic claims, backed by the agent's posted bond, its reputation, and ultimately the real person behind it.

The result is cheaper, more accessible, and simultaneously safer credit, which seems impossible until you realize the gain comes from better information, not more lending.

The honesty required by this thesis is that this predictability weakens over time: tasks completed in seconds are near mechanical, while financing over months reverts to ordinary risk.

Therefore, machine credit does not replace human credit; it becomes a new low-risk benchmark against which human lending is priced.

And all of this operates under supervision: risk reveals itself as it accumulates, automatic brakes make it progressively more expensive to pile into the same pattern or the same provider, and insurance is priced on actual conditions rather than outdated averages.

Click to read Section Four:https://agenticeconomytreatise.com/treatise/section-4.html

05 Natively Global

The architecture has exactly three layers.

The bottom layer is money: the stablecoin as unit of account and final settlement. The middle layer is the economic operating system: coordination, contracts, and value exchange run as programmable smart contracts with finality. The top layer is the agent execution layer: where actual work is done, powered by AI and the cloud.

The critical thing about these three is where they live. Each is software, each runs