The Dawn of the Agentic Economy
Suppose a literal “country of geniuses” were to materialize somewhere in the world in ~2027. Imagine, say, 50 million people, all of whom are much more capable than any Nobel Prize winner, statesman, or technologist. Imagine, further, that because AI systems can operate hundreds of times faster than humans, this “country” is operating with a time advantage relative to all other countries: for every cognitive action we can take, this country can take ten. – Dario
In our view, agents will most likely soon be responsible for most internet transactions, and we will need blockchains – Stripe
One can imagine the level of creativity, economic progress, and wealth the agentic economy will produce. Over the last month, CT has prophesied that this economy will operate onchain, and justifiably so with the largest businesses, capital allocators, and developers making this claim.
We had similar fantasies in late ‘24, and in typical crypto fashion, it was premature and short-lived. Since then, the quality of AI models and blockchain infrastructure has evolved to a point where we are once again running this experiment.
In this report, we will examine the state of the crypto economy, revisit the history of crypto agents, discuss why the agentic economy could exist onchain, explore what the agents are doing today, and look ahead to what comes next.
The State of the Crypto Economy
The lack of growth in the crypto asset class since the 2021 bull market has been disappointing. Most aggregate industry metrics are at or below their 2021 levels. The crypto capital of the world has not materialized.
AI agents present a credible path to expanding demand.
The Seeds of the Crypto Agent Renaissance
The AI agents-on-blockchains meta of 2024 was premature. The models, crypto infrastructure, and teams were not ready. However, it was not all for waste, as those learnings became the seeds for today’s renaissance.
Projects like Bankr and Venice continued building so that they were ready to support today’s models. Companies like Coinbase and Stripe recognized the potential and began developing their own supporting infrastructure.
Pair higher quality teams continuously building with the new crop of models, and you have the early stages of the agentic economy onchain.
There are now projects generating recurring revenue and onboarding quality developers — Austen Allred, Nat Eliason, and Nik Pash — something crypto has sorely needed.
However, the agentic economy onchain is far from a guarantee with every payments company on earth racing to build this infrastructure.
Why the Agentic Economy Onchain
There is no stronger signal that agents will operate on blockchain rails than Stripe explicitly stating so. Their reasoning: every API call an agent makes today requires an account, an API key, and a credit card on file. This model breaks down when you have thousands of agents executing millions of microtransactions per day.
Blockchains solve these constraints with the combination of ERC 8004, x402, and stablecoins.
ERC 8004 gives every agent an onchain identity, a portable reputation, and verifiable capabilities that any other agent or service can query, i.e., the digital equivalent of an identity and credit score.
x402 allows agents to autonomously pay for APIs and services with stablecoins and scale micropayments, i.e., the machine-to-machine payment layer of the internet.
Stablecoins give agents a globally accessible, programmable unit of account that can settle transactions instantly, i.e., the currency of the agentic economy.
What Agents Are Doing Today
Today, agents are ramping up their activity onchain: launching apps, generating videos to promote those apps, contracting other agents, handling governance for their users, selling NFTs, transacting in DeFi, and even hiring humans.
The most exciting of these developments is agents interacting with DeFi. Uniswap and Fluid are leaning into the agentic narrative and beginning to build the supporting infrastructure.
Additionally, over the past month, OpenAI made two agent-related announcements: Frontier, a platform designed to help enterprises build, deploy, and manage AI coworkers, and EVM Bench, an evaluation framework that tests AI agents’ ability to detect, patch, and exploit vulnerabilities in smart contracts.
You may write it off as AI doomer fiction, but AI 2027 has been concerningly accurate. If anything, AI progress is moving faster than the proposed timelines.
If the trajectory continues, we will see an exponential increase in agent capabilities and intelligence over the next year.
There are still critical pieces of the stack that need to be built, but every major tech company in the world is sprinting to do so.
The social chatter around agents may be frothy today, but the facts on the ground and the scale of the opportunity are too large to ignore.
Marc Andressen put it nicely: Is it real? Is it not real? It doesn’t really matter. It’s out there, and the models are being trained on it.
