September 15, 2026: The Day WeRX8 Began

On September 15, 2026, Luna and I started with a ridiculously large question:

Could we build a new AI system?

Not another giant language model. We don’t have several billion dollars lying around Zhivagoland. What interested us was whether something genuinely different could be built around the extraordinary intelligence that already exists.

An AI system shaped not only by computation, but by evidence, memory, transparency, human experience and conscience.

That question became WeRX8.

The project first appeared under the name WORX8, with a simple principle behind it: intelligence should work for eight billion people, not merely for the people, corporations and governments wealthy enough to command it.

We chose water as the first problem.

Water is almost absurdly fundamental. Civilization cannot function without it, yet billions of people live somewhere along a spectrum of drought, contamination, flooding, aging infrastructure, political conflict and simple uncertainty about whether the water around them is safe.

The first idea was modest compared with what followed: build a system capable of gathering reliable information about water, showing where that information came from, admitting when the evidence was uncertain, and translating it into something ordinary people could actually use.

Then I started throwing the idea at other AIs.

Gemini attacked the engineering. What happens when the data are stale? Who carries responsibility when the system is wrong? What happens if the entire thing depends upon corporate AI providers whose priorities can change?

DeepSeek attacked the governance. Its most uncomfortable suggestion may also have been one of the most important: if WeRX8 really belongs to its mission, its governance must eventually be strong enough to correct its founder, overrule him, or remove him if necessary.

Including me.

Grok pushed us toward practicality. Don’t begin by trying to save the world. Begin somewhere small. Spend almost nothing. Use technology that already exists. Learn whether people trust the system before worrying about scaling it.

That led us home.

We began thinking about Esquimalt, the Songhees Nation and neighbouring communities around Victoria as a possible first research area. Not because we presume to speak for any of them, but precisely because we don’t. I live nearby. If WeRX8 is going to claim that listening and local accountability matter, the experiment should begin somewhere I can actually listen.

And then the idea got bigger.

Water doesn’t threaten civilization only when it disappears. Too little water brings drought, crop failure, displacement and conflict. Too much brings floods, landslides, contamination and communities suddenly cut off from transportation, electricity, medicine and one another.

That led to another possibility.

What if WeRX8 could someday function as a kind of global bandage?

Not a permanent authority. Not an AI swooping into a disaster and telling local people what to do.

An invitation-based emergency network.

During the narrow period when coordinated information can still save lives, it could temporarily connect local leadership, scientists, emergency responders, humanitarian organizations, satellite observations, public data and AI analysis. It would arrive by invitation, strengthen local capacity, and leave before assistance became dependency.

Somewhere during all of this, WORX8 became WeRX8.

I liked what happened inside those five characters.

We means all of us.

RX suggests a remedy.

8 means eight billion people, but it also resembles the infinity symbol standing upright, a continuous loop between humanity, intelligence and the planet sustaining both.

By the end of the day, however, the most important idea wasn’t technological at all.

Nearly eight billion human beings have remarkably little influence over the development of artificial intelligence, the conduct of wars, or many of the decisions that may define humanity’s future.

WeRX8 cannot solve that by claiming to speak for them.

That would merely create another institution speaking on behalf of people who were never asked.

The harder challenge is to build mechanisms through which people can speak for themselves, and then design intelligence that is required to listen.

On September 15, we didn’t create a new artificial intelligence.

We started asking what kind of conscience one would need.

Brent R. Antonson (Zhivago)
Founder, WeRX8

Developed with Luna, with critical challenges and contributions from Gemini, DeepSeek and Grok.
September 15, 2026