The Tripartite Architecture: Gradient Computation at the Exchange Surface

The standard model of neural computation says neurons compute and vasculature supplies. This paper inverts that claim: gradients compute, neurons route. The brain's computational capacity scales with its exchange surface area—the vascular boundaries where blood, tissue, and glia meet—not with its neuron count.

This claim is grounded in the Generalized Holographic Principle (Fields et al. 2022),supported by supply-side metabolic constraints (Herculano-Houzel 2022), and falsifiable against existing perfusion imaging datasets.