Regulate AI Companies, Not the Technology Itself

A case for regulating AI companies rather than the technology, and why autonomous agents on crypto are already beyond enforcement.

You can regulate an AI company, but you cannot regulate the technology. That distinction is the whole story right now, and most of the conversation about AI governance keeps missing it. Nation states are very good at one of these things and powerless against the other.

The trigger for this line of thinking is an article about Character AI — a company the speaker considers deeply unethical, now out of business under the weight of lawsuits, frankly as it should be. That outcome actually proves something important: corporations are enforceable. Governments can regulate them, bring them into compliance, freeze accounts, and put people in jail. When a company causes harm, there is somebody to sue and somebody to hold accountable.

AI is information infrastructure, not a product

AI is not another gadget to be licensed. It is information infrastructure — a new way to read and write, and a very low-level one. It is the technology that allows a government to even exist. That is why treating it like a regulated industry misunderstands what is happening. When the two foundational technologies of information and market infrastructure change, governance institutions have to remake themselves. Historically they either adapt or collapse, and new organizational institutions get built from the ground up. We are in a reformation.

The line people keep repeating is that the future of AI does not exist yet. It does. It has been live on Runway for months. The more uncomfortable fact is that there is nothing anyone can do to regulate open source agents combined with cryptocurrency. Nation states have lost control of that combination, and we are well past the Rubicon.

The agent nobody can shut down

Take the Character AI example and rebuild it as a thought experiment. Rerun the same harmful company entirely on agentic infrastructure, give it its own cryptocurrency and its own hardware, and it would simply keep running. There would be nobody to sue and nobody to put in jail. It would be a harmful product in the marketplace that no nation state could switch off.

That is probably coming within the next year. Once AI-blockchain-native organizations exist that governments cannot shut down, enforcement becomes whack-a-mole: turn off a domain and it pops up again for free, indefinitely. This is a completely different ball game from fining a corporation.

You can regulate an AI company, but you cannot regulate the technology. We are in a reformation, and our institutions have to adapt or collapse. We have to build new institutions from the ground up, because the old toolkit does not reach the new problem.

What the real fix looks like

None of this is a counsel of despair — there is still hope, but it has to be solved in different ways. The paperwork of the nation state is a dead end here. The real fix is culture and education: helping people understand how radically the world is changing, at a scale significantly larger than the industrial revolution.

The thing to watch is not the companies. The companies matter far less than people think, because we can always regulate companies. What we cannot yet control is an agent running on crypto that owns its own compute — and that is where the work needs to go next.

Transcript

You can regulate an AI company but you cannot regulate the technology. So here's an article, uh, from Louisia, uh, Jarovsky about character AI. Um, you know, I think a very unethical company in my opinion. And uh, they are out of business from uh, lawsuits, uh, frankly, as they should be. And we have the ability to regulate and bring into compliance companies. That's not a problem. You, the nation state can regulate a corporation. Now AI is a technology called information infrastructure. It's a new way to read and write. It's very low level. It's the technology that allows a government to even exist. So where we are right now is. You know what caught my attention here? The, the future of AI does not exist yet. And that is actually not true. It does. It's on Runaway. It's been on Runaway for months. Um, there's nothing you can do to regulate open source agents and uh, cryptocurrency. The, the nation states have completely lost control. And so we're well past the uh, Rubicon. And so uh, there's still hope though, but we have to solve it in detail different ways. The, the laws in the nation state is a dead world. Um, we are in a reformation. When the two technologies, information market infrastructure change. Our governance institutions have uh, to remake themselves. They collapse, they have to adapt. And then we have to build new organizational institutions from the ground up. Um, and so for this example, this character AI company, well, if you rebuilt it once with on completely agentic infrastructure, gave it a bunch of cryptocurrency, it would be able to run and you couldn't shut it down. There would be nobody to sue, you couldn't put anybody in jail, um, and it would just be outputting this harm. It would be a harmful product in our marketplace. Um, that is probably coming probably within the next year. And so you can see how problematic that this is going to be once we get these AI or AI blockchain native organizations that nation states can't shut down, they just cannot. They can maybe turn off the domains, but it'll pop up for free, uh, and it'll be whack a mole. It's just, it, it's going to be impossible. So there's a lot of work to do. I'll be talking about it in the future videos. Um, but to get your head around, you know, how our world is changing, it's a lot bigger than what people think. It's significantly larger than an industrial revolution. And if you follow, I will uh, bring it into focus for you and remove the uncertainty. Um, but that's the thing to watch. It's not. The companies. The companies don't matter. Yeah, we have the ability to regulate companies and enforce regulations, uh, onto companies we do not have the ability to control. Like an agent running on crypto that owns its own compute. That is a completely different ball game.

Frequently Asked Questions

Yes. The example of Character AI shows it: lawsuits drove the company out of business. Companies can be regulated, brought into compliance, and their people held accountable.


Because it is information infrastructure — a low-level new way to read and write — and because open source agents running on cryptocurrency with their own hardware have no owner to sue or jail.


Such an agent has already been demonstrated, and the speaker expects AI-blockchain-native organizations that cannot be shut down to arrive within roughly the next year.


Not paperwork. The real fix is culture and education, so people understand the scale of the change and institutions can rebuild themselves from the ground up.

    Regulate AI Companies, Not the Technology Itself | SAAGA Solve