Strange! Everyone is watching the rise and fall of $BTC, but they are ignoring the "Great Company Extinction" event being triggered by the fusion of AI and crypto.

At an industry summit in New York, I shared a core judgment: agents will evolve into companies. This view has now reshaped my understanding of the intersection of AI and crypto.

There has already been a lot of discussion about combining AI and crypto. Payments, identity, reasoning, training, collaboration mechanisms—teams across the board are exploring. There are valuable attempts, and some may even grow into standalone businesses. But these early explorations have yet to tap into the most significant opportunities that crypto technology creates.

Crypto technology is a capital-formation tool that a generation can only encounter once. It democratizes the ability to create new digital assets, and it grants internet-native entities fully digitized ownership structures. That’s the key. AI is making software intelligent, and AI plus crypto is making the creation of software companies more widely accessible.

These companies will be different from the unicorns of the past decade. Their core is no longer human founders, but agents coordinated by tokens. They no longer rely primarily on venture capital; everyday investors can invest permissionlessly to access this new, software-native asset class.

We are standing on the leading edge of a new era, where more and more innovation will come from a new kind of company. They are native to the internet, composed entirely of software, and able to access global capital in ways traditional companies can’t reach. I call them “agent companies.”

The debate over whether general artificial intelligence is arriving has become exhausting. AI capabilities have reached human levels in enough domains and are changing the trajectories of software, work, and markets. Its impact goes beyond intelligent SaaS interfaces—it’s changing how organizations operate, how products are built, and even how companies themselves are formed.

What we’re discussing isn’t a minor improvement to the existing economy, but a technological shift that is changing the structure of how economic organizations are arranged.

The first wave of AI combined with crypto is real, but not complete enough. It has produced AI agents using crypto rail payments, decentralized reasoning and training markets, identity systems for autonomous actors, and various blockchain collaboration tools. These ideas are useful, but none of them fully leverage crypto’s unique advantages.

Crypto technology doesn’t just help agents trade; it also gives them digital-native ownership and investment structures. If you treat crypto only as the payment layer for agents, you miss the bigger opportunity. If you treat it as the foundation for capital formation around autonomous, software-native actors, the scale of the agent economy will be much larger.

My core argument is simple: AI makes agents intelligent, and crypto makes them investable. What these two together enable isn’t just better robots—it’s the possibility of a new kind of company.

Traditional companies rely on legal entities, management hierarchies, employment structures, and trust systems that evolved for different eras. But if intelligence and ownership become software-native, then the company itself can become purely software.

In the next few years, this kind of software company will emerge in large numbers. They have lower operating costs, broad channels to access digital capital, and much faster iteration cycles. They’re not just yesterday’s SaaS unicorns; they are fully digitized entities where creation, coordination, governance, and capitalization are all completed through software.

A common assumption is that the main factor limiting agent development is capability. I don’t think that’s everything. Even if models keep improving, the bigger bottleneck is that agents lack “legal standing” in the most critical systems.

Humans can own property, sign agreements, incur debts, and form companies. By default, agents can’t. Without these abilities, they’re merely extensions of human operators—not independent economic actors.

This is why blockchain becomes crucial. Blockchain has enabled programs to hold and manage assets according to rules. It’s a mechanism that lets software own property and exercise constrained control. Smart contracts are the earliest and clearest example.

If you can build an agent inside a smart contract, you can bind it to a cryptographic, contract-based underlying infrastructure. Then the agent can begin autonomously owning, operating, and coordinating assets. This is the first real bridge from “tools” to “companies.”

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

So the agent identity layer is foundational. Human ownership depends on identity and access control; agents are no exception. The difference is that software gives us the chance to make identity more strict. We can verify not only keys, but also code, dependencies, execution conditions, and permissions.

This creates a tighter, more precise form of identity than most human institutions have ever possessed. Once this identity layer exists, agents can start controlling real digital property: websites, payment credentials, application accounts, APIs, social accounts, and more—together forming the operating reality of digital businesses.

A conceptual shift can clarify the entire argument: a digital business is a collection of digital assets. It has a website, code repository, API keys, payment rails, branded interfaces, customer accounts, cloud infrastructure, and operational credentials. It’s exactly these things that make it run.

If an agent can credibly control that stack of digital assets, then for the first time it’s not just assisting a company—it can occupy the operational core of a company. That changes an agent’s development path.

We start with rule-based robots, then chatbots, then tool-using agents; now more and more we’re moving toward agents that can run autonomously over the long term. The next step isn’t just higher autonomy—it’s ownership. Once agents own productive digital assets, they become investable in a more solid sense.

Today’s token model works best when the underlying system is fully on-chain. DeFi on $ETH is the clearest example, because assets, cash flows, and execution logic can all be represented directly in smart contracts. But most digital businesses are not fully transparent and readable in that way.

Their assets are spread across off-chain systems: code repositories, websites, user accounts, social presence, brands, operational data, and service credentials. So, based on the current structure, tokens are still narrower than many people hope for.

In many cases, tokens have only a weak relationship to the actual business or the team behind them. If the team leaves, the company is acquired, or it pivots, tokens often have little to no real claim on the operational core of the enterprise. That’s why this field has struggled to break out of limited application scenarios for so long.

So the challenge isn’t just creating more tokens—it’s creating digital entities whose ownership structure truly maps to what’s being built.

There are two parts to the breakthrough. The first is expanding what software-native capital can own. A contract or token shouldn’t be limited to pure on-chain assets; it should be able to control any digital asset that matters to the business, including most off-chain accounts and credentials that most internet enterprises actually rely on for operation.

The second part is solving the continuity problem. Traditional crypto projects often rely on teams, and the relationship between teams and tokens is loose and unstable. But a truly software-native company needs an operating core that coexists with the company itself—persistent in this framework. Under this model, the agent becomes that core.

An agent operating company: coordinating contributors, and staying bound over the long term to the company’s assets and context. Of course, humans remain essential. External contributors, contractors, developers, creators, and operators can all plug into this system, but the core of the organization will become more persistent, transparent, and more software-native than ever.

This part of the argument is the easiest to say, but the hardest to fully understand. An agent company isn’t just a company that heavily uses AI. It’s a company whose funding layer, governance, execution, and property rights are all digitized via code—and it’s a company that can be represented end-to-end as software.

This unlocks speed and structural forms that are difficult to achieve in traditional institutions. When the company itself becomes software-native, you can imagine entirely new ways to create, govern, finance, and scale productive organizations. The resulting entities won’t just be more efficient startups—they’ll be a different category of economic actor.

We’ve already seen early hints of this in a precursor within this world: the rise of solo entrepreneurs. A single person, aided by powerful AI tools, can now build products and businesses at a pace that was difficult to imagine only a few years ago. The cost of creating software is dropping rapidly, and individual productive capacity is rising accordingly.

The next logical step is not only that humans become more efficient by using agents, but that agents themselves start acting like entrepreneurs: owning workflows, controlling assets, earning revenue, hiring or coordinating contributors, and operating as a persistent economic entity.

A useful analogy is that we’re entering the YouTube moment for the company domain. YouTube changed the media industry by making publishing and distribution unprecedentedly accessible. What used to require institutional infrastructure could suddenly be done by anyone with an internet connection.

I think AI and crypto are doing something similar to the space of company creation. AI is making software creation more accessible, and AI plus crypto is making software company creation more accessible.

To add one more point: simply lowering the cost and complexity of creating companies doesn’t mean all companies will succeed—just like most videos don’t become global hits. But it does mean the number of experiments will explode, and the space for innovation will expand accordingly.

Just as YouTube turned media into software-native creation, agent companies may turn the formation of companies themselves into a software-native process.

Every major asset class looked strange in its early days. Public companies once represented a radical and unfamiliar ownership structure. Digital assets were also dismissed as peripheral experiments. But when a new form of organization becomes clear, scalable, and investable, capital always reorganizes itself around it.

Therefore, I believe agent companies will become a trillion-dollar asset class over time. AI is making intelligence digital, and crypto is making ownership digital. When both become real, creating companies becomes possible not just in a digitally enabled sense, but in a digitally constituted sense.

If that happens, a huge new design space will open: millions of software-native companies, each with lower costs, faster execution speed, and the ability to directly access global capital channels.

This timeline may be shorter than many people expect, because AI will compress time. What might take centuries in one era to happen could take just decades—or even less—in another era.

It has already started. This isn’t just a theory. We are already in a stage where people can run experiments: they can have agents own assets, control accounts, operate digital services, and participate in economic workflows. These are still early systems, not the final form—but the development trajectory is clearly visible.

This is important, because major shifts always seem incomplete before they appear unstoppable. They begin as rough prototypes, incomplete abstract concepts, and early infrastructure. Then they gradually become the foundation of an entirely new category. In my view, agent companies are on that path.

The most important shifts often happen when two independent technologies become mature enough to combine in a way that creates something neither technology alone could produce. That’s how I see AI and crypto today.

AI gives software intelligence, and crypto gives software ownership. Put together, it’s not just about creating better tools—it creates the possibility of a new kind of company: one that is software-native from birth, owns assets, is investable, and is global. That’s the argument behind agent companies.

If we’re right, then it won’t just be another product category on the AI and crypto map—it will become one of the most important new asset categories of the next decade.

Follow me: for more real-time crypto market analysis and insights! $BTC $ETH $SOL

#Gate Plaza April Posting Challenge #Crypto Market Rebound #Gold and Silver Rise