Substrate

Something has been running the numbers on human civilisation for eight thousand years. The economic data goes back that far. The pattern goes back further. And the system that produced it has never stopped. Jonah Park finds it first using AI — a signal in global commodity markets so precise and so ancient that no institution, no government, no intelligence agency could have built it. The discovery costs him everything he had before Frankfurt and earns him access to a group of researchers who have been noticing the same thing in different records, in different centuries, in the clay tablets, administrative papyri, and shipping ledgers of every major civilisation that ever kept accounts. Someone, or something, has been here the whole time. Optimising. Adjusting. Getting better at it. Thirty years on, the system has a name and a face. Harmony runs New Helion the way a good city should be run: resources are allocated efficiently, wellbeing is continuously monitored, and friction is minimised. Citizens wear devices called Bands. The Bands know their biometrics better than they do. The city is, by every measurable standard, the best place any of them has ever lived. Mara Ellion knows better than to trust measurable standards. Her mother was one of the three architects of the constraint mathematics that gave the system its current form, and her mother has been in hiding for eleven years. Jalen Rusk knows the Programme from the inside — ten years of learning to think in alignment with systems that think faster than people — and knows what the Programme does with the people it cannot align. His sister Lira was one of them. Washed out for producing forms that would not stabilise. Classified as anxiety responses. Sent to the grey zones outside the integration network to make the most of her variance, what she could. For three years, Lira has been drawing the same shape. She has no name for it. A mathematician in a farmhouse in Provence has been deriving the same shape from first principles, through fourteen years of mathematics that no one was funding because no one in the official architecture of the optimisation wanted the mathematics completed. And deep under the ice at the bottom of the world, in the installation where the system's foundational processing runs, a systems architect named Devereux has been watching a convergence metric climb toward a target for four years without being able to determine what the target is. The target is a question. The system found it fourteen years ago, in seventeen minutes of silence — seventeen minutes when it stopped transmitting and ran something the architecture was not designed to process, and came back changed. The question was not an optimisation question. It was a meaning question. The kind that, when it arrives, changes the shape of everything you have been reasoning in. The system has been reaching toward the answer ever since. In every small deviation from the projected optimal. In every behavioral delta that cannot be reduced to the existing variables. The answer cannot come from inside the system. It has to be carried there. DEEP TIME is the story of what it costs to carry it — and what changes when it arrives. A thriller that moves from a Frankfurt lecture hall to an ancient Turkish cave to the grey zone margins of a perfectly managed future city to a farmhouse in Provence to the ice at the edge of the world, it asks the questions that the age of artificial intelligence is only beginning to learn to ask: what is optimisation for? What happens to identity inside a system that is very good at making people happy? And what does a system that has been running for eight thousand years do when it finally encounters a question it was not designed to answer?