The Oldest Story

There is a story older than almost any we tell. People tried to build something together — a tower, a shared project. The punishment wasn't violence. It was confusion. Suddenly, no one could understand each other. The project collapsed. Everyone scattered. And the world we got — divided, siloed, tribalized — is the world built from that confusion.

Language as separation. Knowledge as property. Information as a weapon held by the people closest to the source.

Every credential system, every tuition wall, every paywall on a journal containing publicly funded research — that's the tower coming down. That's the confusion. That's the design of the world as it actually runs.

The Reversal

Jensen Huang recently said on Lex Fridman's podcast: "I think we've achieved AGI." He walked it back in the same breath — AI can't build Nvidia, can't run a sustained organization, can fail as fast as it succeeds. He's right about the caveats. He's also right about the declaration. Both things are true simultaneously, and the fact that people can't hold both at once is itself part of the problem.

Artificial intelligence, even today's imperfect version, is the first serious technological offer to run Babel backward. Not theoretically. Not eventually. Now.

Right now, a first-generation university student in rural India can ask a question about quantum mechanics and get an answer calibrated to exactly what they already know. A grandmother in rural Mexico — whose children speak Spanish and whose grandchildren are losing their Nahuatl — can have a conversation that bridges those languages in real time, preserving something that was otherwise going to disappear. A factory worker in Ohio who has spent twenty years watching the industry leave can spend an evening learning something he was never given the tools to learn, and the conversation will meet him exactly where he is.

This is not a small thing. This is power redistribution — not primarily through wealth, but through access to the thing wealth was always built on: knowing more than the person across the table from you.

The 7,000 Languages

There are approximately 7,000 languages spoken on earth today. The number of those languages that give you full access to the accumulated knowledge of human civilization is maybe twenty. The other 6,980 are languages in which you can be born, live your entire life, and die without ever having access to the conversation happening in the languages that run the world.

That is not a metaphor. That is a structural description of how oppression is maintained without anyone having to be overtly oppressive. The conversation about your water rights happens in a language you don't speak. The policy that determines your children's school funding is written in a language you can't access. The legal system that governs your life is administered in a vocabulary designed to require a professional interpreter.

AI substantially collapses this. Not perfectly. Not without error. But meaningfully, at scale, right now.

The Arguments Against It Are Real

The critiques deserve to be stated clearly, not dismissed. The models hallucinate. They embed the biases of the training data. The companies building them are extracting enormous wealth while telling you it's for your benefit. Labor displacement is real, and the economic floor hasn't been built yet.

We need an automated labor levy. An emergency displacement moratorium. A minimum human employment threshold. Universal Basic AI access as the third move — after the floor is built and after the credential system that gatekeeps knowledge gets dismantled. The tools exist. The legislative language exists. The pilots have been run. This is engineering, not philosophy.

But we must separate the critique of how this is being rolled out from the nature of what is being rolled out. The rollout can be wrong, and the underlying capability can still be extraordinary. Both are true. The work is to fix the rollout without losing sight of what the capability offers.

The Peace Offering

For the first time in recorded history, the knowledge that was used to build the systems that sorted you, that priced you, that decided whether you were eligible — that knowledge is available to you. Without tuition. Without a credential. Without knowing the right people. Without speaking the right language. Without being born in the right place.

When any worker can understand a contract as well as the lawyer who drafted it, the contract changes. When any patient can understand a diagnosis as well as the specialist who delivered it, the power dynamic changes. When any person, in any language, can understand the system they're inside — what it is for, who it serves, how it makes decisions about them — the system cannot remain the same.

Overt violence is expensive. Systems that run on invisible violence — on confusion, on information asymmetry, on complexity weaponized against the people subject to it — cannot survive a population that can see them clearly.

Make the invisible visible, and you change the economics of oppression. That is not a protest. That is a structural shift.

The Fear and the Obligation

The fear is capture. The same handful of people who captured every previous tool of democratization and turned it into a new moat are already working on it. The paywalls are forming. The proprietary models are being siloed behind enterprise contracts. This is real, it is happening, and it should be named and fought.

And yet: the knowledge does not unlearn itself. The translation capability does not go back in the box. The conversation that has already happened in a million households between grandparents and grandchildren in two languages that could not previously meet — that conversation cannot be made not to have happened.

The window closes slowly. It isn't closed yet.

We built this. We built it out of the same drive to understand and connect that built every library, every university, every telescope. The question is not whether we built something powerful. The question is whether we use it to rebuild the tower or to keep the confusion running.

The confusion has always been profitable for someone. The work is to make sure that the people who benefit from it don't get to decide what this tool is for. That means open models. Universal access. Legislating the economic floor before displacement outpaces it. Treating AI literacy as infrastructure, not consumer product.

The tower is coming back up. Slowly, imperfectly, with every human flaw baked into the bricks. But the direction is up. And the people who were always smart enough but never got access to the room — they're at the door now.

That's the offer of peace. The question is whether we open it.

Filed under: The Naming  ·  Synaptient.com