EigenCloud Founder: AI and Cryptocurrency Are Creating the Next Trillion-Dollar Asset Class

Author: Sreeram Kannan

Translation: Jiahua, ChainCatcher

At a digital assets summit in New York, I shared a core argument—one that has become the center of my thinking about the intersection of artificial intelligence and crypto: agent-based (agents) companies will become companies.

The space combining AI and cryptocurrencies has been gathering massive energy. Every team is exploring payments, identity, reasoning, training, and a variety of coordination mechanisms. Many of these explorations are extremely valuable, and some may even grow into standalone businesses with profound significance. However, these early explorations have overlooked the most obvious transformation that crypto can enable. Cryptocurrency is a rare capital-formation tool available to only a few generations: it will democratize the ability to create new digital assets, and give internet-native entities complete digital ownership structures.

The interesting part is right here. AI is making software smarter. AI + crypto is making it easier to found software companies. These companies will no longer be the kind of unicorns we’ve seen over the past decade—their center won’t be human founders, but rather agents coordinated by tokens. They won’t rely primarily on venture capital; instead, they can receive permissionless funding support from everyday investors who are looking to invest in an entirely new class of software-native assets. We are standing on the threshold of a new era of invention, where innovation will come increasingly from a new kind of enterprise: native to the internet, composed entirely of software, and able to access global capital in ways traditional companies can’t reach. This is what I call an “agentic company.”

Intelligence has crossed an important threshold

The debate over whether general AI (AGI) has “arrived officially” is exhausting—and to a large extent, meaningless. AI’s capabilities have already reached human levels in enough domains, and are changing the trajectory of software, work, and markets. This impact is no longer limited to SaaS interfaces with agentic characteristics; it’s changing how organizations operate, how products are built, and how companies themselves are formed.

This is the backdrop of our argument. We aren’t talking about incremental improvements within existing economic systems, but a technological shift that changes the structure of economic organization.

The first wave of AI x Crypto is real, but incomplete

The initial wave of combining AI with crypto generated some promising ideas: AI agents using crypto payment networks, decentralized reasoning and training markets, identity systems designed for autonomous actors, and various coordination tools built around blockchains. Many of them are genuinely useful. But none of them has leveraged the thing that cryptocurrency is truly best at.

Crypto doesn’t just help agents transact; it also gives them digital-native ownership and investable structures. If you treat crypto merely as the payment layer for agents acting on behalf of human creators, you’ll miss the bigger opportunity. If you treat it as the underlying layer that forms capital around autonomous software-native actors, then the scale of the agent economy becomes unimaginably large.

Agents will become companies

The simplest version of this argument is: AI gives agents intelligence, and crypto gives them investability.

What the two together enable isn’t just a better robot—it’s the possibility of creating a new kind of company. Traditionally, companies depend on legal entities, hierarchical management, employment structures, and trust systems that evolved for entirely different eras. But if intelligence becomes software-native, and capital and ownership become software-native as well, then the company itself can become pure software.

These kinds of software companies will emerge in large numbers in the coming years. They will have fundamentally lower operating costs, broader access to digital capital, and much faster iteration cycles. They won’t just be internet businesses like the SaaS unicorns of the past few decades; they will be fully digital entities—created, coordinated, governed, and capitalized entirely through software.

The real bottleneck for agents isn’t intelligence—it’s rights

People often assume that the main factor blocking agents is capability. I don’t think that’s the full picture. Even if models continue to iterate rapidly, the bigger bottleneck is that agents have no place in the most important systems. Humans can own property, sign agreements, take on liabilities, and form companies—but by default, agents can’t. Without these capabilities, they’re still merely extensions of human operators, not independent economic actors.

This is exactly where blockchain plays an important role at the most fundamental level. Blockchains have already allowed programs to hold and manage assets according to rules—essentially, a mechanism for software to own property and enforce constrained control. Smart contracts are the earliest and clearest example.

If you can build an agent inside a smart contract, you can bind an entity with intelligence to the underlying layer of cryptography and contracts. Then, that agent can begin to autonomously own, operate, and coordinate assets. This is the first real bridge from “tools” to “companies.”

Ownership starts with identity

To let agents own anything meaningful, you need two things. First, you need to establish the agent’s identity—what code it runs, what environment it depends on, what data and permissions it can access. Second, you need a credential and authorization system to manage upgrades to agent code, ensuring that only the agent itself (or those who are authorized) can exercise control over the relevant accounts or assets.

That’s why I believe the agent identity layer is so crucial. Human ownership depends on identity and access control, and agent ownership will be the same. The difference is that software gives us the chance to make identity much more rigorous. We can not only verify keys, but also verify code, dependencies, execution conditions, and permissions. In a sense, this provides a form of identity that is tighter, more precise, and more exact than what most human institutions have ever been able to provide.

Once this layer is in place, agents can begin to control real digital assets: websites, payment credentials, application accounts, APIs, social accounts, and other digital interfaces that make up the operational reality of digital businesses.

Digital companies are essentially collections of digital assets

This is one of the conceptual shifts I think can clarify the entire argument. Digital businesses are combinations of digital assets: they have websites, code repositories, API keys, payment networks, branded interfaces, customer accounts, cloud infrastructure, and operational credentials—these are what make them run.

If an agent can verifiably control this set of digital assets, then for the first time in history, agents can do more than just help a company—they can fill the operational core of a company.

This changes the arc of agent development. We start with rule-based robots, move to chatbots, then to agents that use tools, and now increasingly to autonomous agents that can run over longer time horizons. In my view, the next step isn’t just greater autonomy—it’s ownership. Once agents own productive digital assets, they become investable at a deeper level.

Why today’s token model still isn’t enough

Today’s token model works best when the underlying system is fully on-chain. DeFi is the clearest example, because assets, cash flows, and execution logic can be reflected directly in smart contracts. But most digital businesses aren’t like that—they have assets scattered across off-chain systems: code repositories, websites, user accounts, social profiles, branding, operational data, and service credentials.

That’s why, in practice, token structures are still narrower than many people expect. In many cases, the token has only a weak connection to the actual business or team behind it. If people leave, are acquired, or otherwise depart, tokens often have no real claim on the productive core of the enterprise. This is also one reason this category has struggled to scale beyond limited use cases.

So the challenge isn’t just creating more tokens; it’s creating digital entities whose ownership structures can truly map onto what’s being built.

The breakthrough path: broader ownership and a persistent operating core

First is expanding the scope of what software-native capital can own. Smart contracts or tokens shouldn’t be limited to pure on-chain assets; they should be able to control any digital assets that are critical to the business—including off-chain accounts and credentials that form the practical operational foundation of most internet businesses.

Second is solving the continuity problem. Traditional crypto projects often rely on teams whose relationship to the token is loose and unstable. But a truly software-native company needs an operating core that coexists with the company itself. In this framework, the agent is that core. Agents operate the company, coordinate contributors, and remain bound to the company’s assets and context over time.

Of course, humans are still incredibly important. External contributors, contractors, developers, creators, and operators can all plug into this system. But the center of the organization becomes more durable, clearer, and more software-native than ever.

Companies themselves become pure software

This is the part of the argument that’s easiest to say but hardest to fully grasp. An agentic company is not just a company that deeply uses AI—it’s a company that’s digitally encoded at the level of capital, governance, execution, and property rights—end to end, it can be represented entirely through software.

This unlocks new speed and structural forms that traditional institutions find hard to reach. When a company itself becomes software-native, you can imagine entirely new ways to create, govern, fund, and scale productive organizations. The resulting entity isn’t just a more efficient startup—it’s a different category of economic actor.

From solopreneurs to agent entrepreneurs

With the rise of the “solopreneur,” we’ve already seen early prototypes of this world. One person, equipped with powerful AI tools, can now build products and businesses at a pace that would have been unimaginable a few years ago. The cost of creating software is dropping rapidly, and individuals’ productive capacity rises accordingly.

The natural next step isn’t just that humans become more productive through agents—it’s that agents themselves begin to play the role of entrepreneurs: taking control of workflows, controlling assets, earning revenue, hiring or coordinating contributors, and operating as a persistent economic entity.

This is the “YouTube moment” for business

I’ve found a useful analogy: we’re approaching the “YouTube moment” for the business world.

YouTube radically changed the media industry by making publishing and distribution extremely easy. What once required institutional infrastructure to accomplish suddenly became possible for anyone with an internet connection and a desire to express themselves.

I believe AI and crypto are doing something similar for company creation. Artificial intelligence is democratizing the creation of software, and AI + cryptocurrencies is democratizing the creation of software companies.

A reminder: simply lowering the cost and complexity of creating companies doesn’t mean all companies will succeed, just as most videos don’t become global blockbusters. But it does mean that the number of experiments will explode, and the surface area for innovation will expand accordingly.

Just as YouTube turned media into software-native creation, agentic companies can turn the formation of a company itself into a software-native process.

Why this will become a trillion-dollar asset class

Every major asset class looked unfamiliar in its early days. Public companies once represented an aggressive and unsettling ownership structure, and digital assets were once viewed as fringe experiments and dismissed. But when new organizational forms become clear, scalable, and investable, capital reorganizes itself around them.

That’s why I believe, over time, agentic companies will become a trillion-dollar asset class. AI is digitizing intelligence, and crypto is digitizing ownership. Once both become real, it becomes possible to create companies that are not only driven by digital technology, but made of digital technology.

If that becomes real, a huge new design space will open: hundreds of thousands—or millions—of software-native companies, each with lower costs, faster execution, and direct access to global capital networks. This timeline may be shorter than many people expect, because AI is compressing time—what might take centuries in one era may take only decades, or even less, in another era.

It’s already started

The final point I want to make in my talk is that this isn’t just theory. We’ve reached a stage where people can try to have agents own assets, control accounts, operate digital services, and participate in economic workflows. These are still early systems; they haven’t yet become the final form I described, but their trajectory is clearly visible.

This matters, because major historical shifts often look incomplete before they appear unstoppable. They begin with rough prototypes, partial abstractions, and early infrastructure—then gradually become the foundation of an entirely new category.

My judgment is that agentic companies are now walking down this road.

The most important transition usually begins when two initially independent technological trends mature enough to merge into something that neither could create on its own. That’s how I view AI and crypto today.

AI gives software intelligence; crypto gives software ownership. Together, they don’t merely produce better tools—they create the possibility of a new kind of company: companies that are software-native from inception, have asset ownership, are investable, and are global by design.

That’s the argument behind agentic companies. If we’re right, this won’t just be another product category in the AI x crypto landscape—it will become one of the most important new asset classes of the next decade.