Population as control
At N=100,000 the facility knows what normal looks like. A bad lot bends thousands of curves at once and points at itself. Nobody reserves material to prove the experiment happened.
biological datacenter program · proposal v0.22
Ask a hundred mice, get a maybe.
Ask 100,000, get an answer.
Biology is a low-resolution science because every datapoint has a price. We are engineering that price to near zero — a vertical facility where a million mice live above a liquid-nitrogen biobank, gravity is the courier, and the population is the control.
the argument
Don't make the cost cheaper. Remove it from the system.
No hands, no fluidics
Zero human touches between birth and biobank. The only liquid handling in the building is performed by the mice. If a step needs a pipette, the step gets redesigned.
Physics instruments only
Gravity, cold, optics, radar, gas sensing, mass spectrometry — instruments whose per-sample cost is fixed, not linear. Reagents only where they train imputation.
the economics
All-in cost per fully-characterized mouse-life
fig 2 · illustrative model · log-log
the build
Three artifacts, one curve
drawing shm-12-001 · rev a
M1 · The Mouse Box
months 0–18 · $1.5–2.5M
A sealed habitat that is itself the instrument: whole-body calorimetry, radar, thermal, self-service dosing, a terminal gravity path into cryostorage. 500 units in the first racks.
proves: the cost curve bends
live twin → drawing pod-40-002 · rev aM2 · The Pod
months 12–36 · $10–20M
Ten thousand Boxes over a vertical LN₂ microbank, pneumatic sample transport, mass-spec lines at 2,000+ samples a day, and the first real cohort: 10,000 mice, one reference compound, every tissue.
proves: cost per sample flat with volume
live twin → drawing bdc-100-003 · conceptM3 · The Datacenter
months 30+ · $100M+ · financed like infrastructure
One million mice in ten cohorts. Compound in, a 14-day whole-body multi-omic movie at n=14,000 out — plus 86,000 monitored survivors, forever. Under 1% of annual global mouse use, more standardized in vivo data than the rest of the world combined.
proves: marginal cost per mouse ≈ 0
walk the building →