PreTect: Thinking Inside the Future

Some ideas don’t arrive fully formed.
They gather slowly, like weather.

PreTect began not as a product, but as a discomfort — a recurring sense that many of the most consequential decisions we make are taken without anywhere proper to stand.

Buying a house is a good example. You can see what a building is now. You can imagine what it might become. But there is often a missing middle: a way to understand whether those two things genuinely connect. Decisions get rushed not because people are careless, but because imagination has no structure. We leap from photographs to fantasies, from surveys to Pinterest boards, without a place to think inside the future we’re contemplating.

Earlier this year, I read a piece in The Economist about a weakness in current AI systems: despite their fluency, they lack what cognitive scientists call a world model. They can describe the world, but they do not inhabit it. Humans, by contrast, carry an internal sense of space, constraint, and consequence. We know — intuitively — that a car won’t suddenly float, that a leaf will eventually hit the ground, that a room has limits even when we close our eyes.

The argument struck me because it articulated something I had been circling from another direction. Much of what feels wrong about contemporary “AI assistance” isn’t intelligence as such — it’s disembodiment. Too many tools generate options without consequence, images without gravity, futures without friction. They narrate possibilities rather than situating us inside them.

World models, as described in that article, are not about spectacle. They are about coherence. About persistence. About the quiet realism that allows you to ask what if without being misled.

PreTect emerged from asking a simple question:
What would it look like to apply that kind of thinking — not to robots or simulations — but to human decisions that carry real weight?

In practice, this meant resisting many of the things that are usually expected. PreTect does not generate endless alternatives. It does not optimise. It does not promise transformation. Instead, it explores a single, plausible future of a specific house — one that takes constraints seriously, that understands scale and setting, and that remains consistent no matter how it’s viewed.

This is not architectural design, and it is not planning advice. It sits deliberately upstream of both. Its purpose is orientation rather than execution: to give someone enough grounded understanding of a future they might live in to decide whether to proceed at all.

At this stage, PreTect is still forming. The language is settling. The boundaries are being tested. What feels clear already, though, is the underlying conviction: that people don’t always need more answers — they need somewhere to think that doesn’t collapse under scrutiny.

If PreTect works, it won’t feel clever. It will feel calm. Like standing in a place long enough for it to tell you whether it’s yours.

The product will come later. For now, this is the idea it’s growing from.