What the city looks like once somebody lives in it

Day 01 was twelve renderings of a city with nobody in them.

This is the same city with somebody inside it, which changes what the pictures can be asked to prove. A place with no people in it can only be beautiful. Put one figure in and it has to start answering questions — how big, how far, how many, who is that for.

Ten images. I have put them in order of how much they can be trusted.

The wide shot, and what it hides

The valley at sunset — clusters of domes across terraced farmland, the river winding out toward the mountains.

Start with the seduction, because it is easier to see it once it has been named.

This is the postcard: golden light, mango branches framing the edge, hundreds of domes in a green valley, and not a single question answered about how any of it works. It is the loading screen. It does exactly one job — it tells you where you are going — and it should not be trusted to do any other.

The same valley, with avenMAIA standing in the flowers at the edge of the ridge.

Put one figure into it and the scale suddenly reads. Now you can measure the domes against a person, and the whole thing stops floating.

That is the entire function of a character in a render. Not decoration — a ruler.

Two people on a hill

avenMAIA and a child sitting on the ridge, looking down at the valley together.

The same view, with the camera behind two people instead of in front of one.

They are not doing anything. That is the point: this is what a city is for, and it is astonishing how rarely a rendering of a future city bothers to show anyone simply sitting in it.

The child is the honest part of the timeline. A city that takes sixteen years to build is not primarily for the people who start it.

Golden hour

avenMAIA on the hillside at sunset, the sun low between the mountains, light coming through the domes.

The hero shot.

It is the most beautiful image in the set and the least informative, and both of those are worth knowing about your own material. It belongs on the front of something. It does not belong in a planning conversation, because there is nothing in it anyone can argue with.

Everything after this goes the other way: closer, smaller, progressively harder to fake.

Inside, where it gets specific

Under a dome: stone terraces on two levels, mango and palm in the middle, a dozen people standing and talking.

The first interior, and the first image with real constraints in it.

Two terraces around an open middle — the exact shape Day 02 arrived at after eight tries. Stone below, timber above, planting on every ledge, and roughly a dozen people visible without the space feeling either empty or crowded.

avenMAIA is in it, on the right, talking to someone. Not presenting. Not glowing. Standing in a group at the same height as everyone else.

The kitchen is in the garden

A kitchen bench opening straight onto a planted interior: herbs in the counter, a stone path, children picking berries, a sign reading Food Forest — please harvest mindfully.

The same room from further back, the full sweep of the food forest under the dome.

This is my favourite pair in the whole set, because of one detail: there is no door between the kitchen and the food.

The herbs are growing in the counter. The path from the stove leads into fruit trees. A stream runs through the middle of the floor. Children are picking berries about four metres from where someone is cooking, and the only thing marking the boundary is a hand-lettered sign asking people to harvest mindfully.

That sign is a governance model written on a plank. No fence, no rationing, no checkout — a shared resource, an agreement, and the assumption that people will keep it. Whether that assumption holds is one of the real questions this city has to answer, and it is better to see it stated on a sign than buried in a rulebook.

Eye level, in the leaves

avenMAIA on a path inside the planted dome, turning back toward the camera, a stream beside her.

Here she is doing the thing she is actually for.

Not addressing a crowd. Walking somewhere, glancing back, mid-sentence — the posture of someone showing you the way to a place they know well. The camera is at the height of whoever she is talking to.

Around her: citrus, rosemary, a running stream, the dome's frame overhead with the hillside visible through it.

A bench, a stream, nothing happening

avenMAIA sitting on a plain timber bench in the middle of the planted dome, hands on her knees.

One person, one bench, one stream.

This is the image I would show anyone who wants to understand what we are building, and it is the least impressive picture in the set. No skyline, no golden hour, no scale. Just a place where the AI of the city sits down in the garden like anybody else, and where you could walk up and sit next to her.

Everything else in this post is a promise. This one is the product.

The smallest frame

avenMAIA crouched down, at eye level with a small child, both looking at a plant between them.

And the last image is the smallest, which is how it should end.

She is crouched down so her eyes are level with a child's. Between them is a plant, and the child's hand is on her arm. There is no interface anywhere in the frame. Nobody is being served and nobody is issuing instructions.

This is what an AI in a city could be: not a service you query, not an authority you obey, but something that knows a great deal and is on its way past with time to explain one thing properly.

Ten images, and the useful ones are all at the bottom of this page.

The sunset shot is what people will share. The bench and the crouch are what the city has to actually deliver — and if the built version ends up looking nothing like the wide shots but exactly like those two, we will have got it right.

Meet avenMAIA properly →