Life imitates art.
When people imagine beautiful futures — and share those visions — the world moves toward them. The stories we tell about tomorrow shape the tomorrow we build.
beacn exists because we believe the most powerful thing you can do for the future is imagine it well.
But there's a problem.
What does a “good” future actually look like?
You might picture a world powered by clean energy. Someone else pictures a world where humanity lives among the stars. Another person imagines a neighborhood where everyone knows each other's names.
Are they all right? Are some better than others? Who decides?
This is the deepest question in philosophy — and now, it's also the deepest question in AI. Because we're building machines that will be extraordinarily good at accomplishing goals. The question is whether those goals will be good.
No single person, no committee, no company can answer that alone. Goodness is too vast, too complex, too personal, too contested.
So beacn lets everyone answer it. Together.
The fastest way in: the Future Compass
You see two visions of the future side by side and pick the one you'd rather live in. Ten choices later, you see where your compass points: the themes your choices lean toward and away from, and how often you sided with everyone else. No account needed.
Each choice is a pairwise preference, the same kind of human signal used to train AI systems with feedback. Choices are anonymous and published as an open dataset so anyone, including the people building AI, can study what futures people actually want.
Find your Future CompassHere's how it works.
You imagine.
Post your vision of the future you want. Write about it. Paint it. Photograph it. Record a short video describing it. Share an article that captures it.
It can be cosmic — how humanity solves its greatest challenges. Or it can be intimate — what dinner looks like when we get the future right. There's no wrong scale. Every vision matters.
The only rule: point forward. Toward something better.
Everyone votes.
Other people see your vision and vote with their values. Upvote means “yes — I want that future.” Downvote means “no — not that one.”
No algorithm decides what's good. No editorial board. No single definition. Just people, honestly responding to each other's visions.
A direction emerges.
This is where it gets extraordinary.
Every vision on beacn exists as a point in a vast space — like a star in a night sky. Visions about similar ideas naturally cluster together. Clean energy near sustainable cities. Space exploration near frontier science. Community gardens near neighborhood design.
When you vote, you're giving a star more gravity. Every upvote makes that vision heavier — it pulls harder. Every downvote takes weight away. The stars with the most gravity are the futures the most people believe in.
Now imagine a beacon in the middle of that sky, being pulled by every star at once. The stars with the most gravity — the visions with the most votes — pull it the hardest. Millions of visions. Millions of votes. And the beacon drifts toward the futures humanity wants most.
We call that direction the goodness direction.
It's not one person's definition of good. It's not a manifesto or a rulebook. It's the living, evolving answer to the question humanity has been asking since the beginning: what should we become?
The constellation
Open the Explore tab and see it for yourself. A sky full of stars. Each one is a vision someone posted.
Stars that are close share similar ideas
You'll see neighborhoods form — clusters about technology, nature, community, art, space, health, education. The map organizes itself by meaning, not by algorithm.
The brightest stars have the most gravity
More votes, more pull. You can see at a glance which visions are pulling hardest on the beacon.
The glowing beacon is the goodness direction
Pulled by every star, drawn toward the futures with the most gravitational force. It's the place in the sky where humanity's collective vision lives.
It's never finished. It shifts as new voices arrive. It's always becoming.
Why this matters now
We are building the most powerful technology in human history. AI systems that will shape healthcare, education, cities, policy, and more. These systems will need to know what “good” means.
Right now, that definition comes from small groups of researchers at a handful of companies. beacn offers something different: a definition of good that comes from everyone. Continuously updated. Collectively authored. Expressed through art, writing, video, and imagination — not just checkboxes on a survey.
Every vision you share and every vote you cast adds to a signal that can actually guide these systems. Not as a wish — as data, in the mathematical language AI already speaks.
Your art shapes the future twice: once by inspiring the people who see it, and once by teaching the machines that will help build it.
For artists
One of the great purposes of art is to elevate the spirit of an idea — to give thoughts and feelings power.
“I want cities with more trees” is a vision of the future. A painting of a city wrapped in forest canopy, light filtering through the leaves, children playing in the shade — that's the same vision, but now you can feel it. Now you want to live there.
Both are art. Both shape the goodness direction. But the painting shows you why art has always been one of the most powerful forces on earth.
beacn exists to give the most power possible to the spirit of a good today and a good tomorrow. Every post — a sentence, a sketch, a photograph, a film — is welcome here. And if you're someone who can make people feel a future, you already know what you can do with a place like this.
We encourage you to elevate your vision however you can. Add an image. Record a video. Write an essay, a poem, a song. The richer your expression, the more fully it captures the future you see — and every form of expression helps shape our shared definition of goodness.
For researchers
beacn produces a novel alignment signal: a continuous vector in multimodal embedding space, derived from collective human voting on future-state visions.
Each post is embedded using Gemini Embedding 2 (text, images, video → unified 3072-dimensional space). User votes define a goodness direction vector: normalize(mean(upvoted) - λ·mean(downvoted)). This is a democratic, continuously-updated, multimodal alternative to RLHF preference data.
Open research questions include convergence properties of the goodness direction, cluster structure of individual preference vectors, and whether this signal captures alignment-relevant information that standard preference data misses.
For everyone
You don't need to be an artist or a researcher. You just need to care about what happens next.
Every vision you post — even a few sentences about the world you want your kids to grow up in — adds to the clearest picture humanity has ever drawn of what “good” means. Every vote you cast gives a star more gravity.
The future isn't something that happens to you. It's something you imagine first.
The beacon needs your light.
It's early. Constellations are forming, but they need more stars. Your vision, your vote, your perspective matters more right now than it will when millions of people are here.
Help us answer the oldest question in the world — and make sure the answer reaches the most powerful technology we've ever built.