Recently, I’ve been pondering a question: if AI can make software smarter, and encryption technology can digitize ownership, what happens when these two come together?

My answer might be surprising — it’s not just a better tool, but a completely new form of company.

You see, many people are now experimenting with AI encryption — payments, identity, reasoning training — these areas all have value. But they’ve missed a more fundamental point: the core capability of encrypted assets isn’t just payments, but enabling software-native entities to own, control, and operate assets. That’s the real breakthrough.

In other words, intelligent agents will become companies.

Imagine: a code-driven intelligent agent, no longer just a tool for a human founder, but a genuine economic entity. It can own assets, sign agreements, control accounts, generate revenue. It sounds like science fiction, but the technical foundation already exists. Smart contracts allow programs to hold and manage assets according to rules, and blockchain provides cryptographic infrastructure. What’s missing now is truly connecting AI’s intelligence with encrypted ownership mechanisms.

But here’s a key bottleneck — it’s not that AI isn’t capable enough, but that intelligent agents currently lack identity and rights within existing systems. Humans can own property, sign contracts, form companies, but by default, intelligent agents cannot. Once this problem is solved, everything changes.

The real breakthrough lies in the identity layer. You need to establish verifiable identities for intelligent agents — what code they run, what environment they depend on, what permissions they have access to. Then, through a combination of cryptography and smart contracts, enable these agents to control digital assets: websites, APIs, payment accounts, social accounts — these are the true backbone of internet business.

This is why today’s token models are still insufficient. DeFi’s success is because assets and execution logic are on-chain. But most digital businesses aren’t like that; their assets are scattered across off-chain systems — codebases, user data, brands, operational credentials. So current tokens often have a loose connection to actual business operations. The real solution is to enable smart contracts and tokens to control off-chain assets while maintaining continuity — with intelligent agents as the operational core of companies, bound to the company’s assets and growth.

When all this becomes reality, what will you see?

Software-native companies. From capital, governance, execution, to ownership, all encoded digitally. These companies will operate at a fraction of the cost of traditional firms, able to directly tap into global capital networks, and iterate much faster. They won’t just be startups deeply leveraging AI, but a completely different economic species.

This reminds me of the impact YouTube had on the media industry. Tasks that once required institutional infrastructure suddenly became accessible to anyone with an internet connection. AI and encrypted assets are doing the same for business creation — democratizing software development, democratizing company formation.

Of course, not every company will succeed, just as not every video becomes a global hit. But that means the number of experiments will explode, and the surface area of innovation will expand accordingly.

Why do I believe this will become a trillion-dollar asset class? Because every major asset class looked unfamiliar at first. Public companies once had radical, unsettling ownership structures; digital assets were once fringe experiments. But when new organizational forms become clear, scalable, and investment-worthy, capital reorganizes.

This timeline might arrive faster than many expect. AI is compressing time — tasks that once took centuries might now take only decades or even less.

Most importantly, it’s already beginning. There are people experimenting now with giving agents ownership of assets, control over accounts, operation of digital services. These are early systems, far from complete, but the trajectory is clear.

Major historical shifts often appear incomplete before they seem unstoppable. They start with rough prototypes and early infrastructure. My judgment is that intelligent-agent companies are now on this path. If we’re right, this won’t just be another product category in the AI encryption landscape, but one of the most important new asset classes of the next decade.