An Uncertainty Principle for Probabilistic Computation in the Retina

TL;DR

None

Contribution

None

Abstract

We introduce a probabilistic model of early visual processing, beginning with the interaction between a light wavefront and the retina. We argue that perception originates not with deterministic transduction, but with probabilistic threshold crossings shaped by quantum photon arrival statistics and biological variability. We formalize this with an uncertainty relation, \( \Delta \alpha \cdot \Delta t \geq \eta \), through the transformation of light into symbolic neural code through the layered retinal architecture. Our model is supported by previous experimental results, which show intrinsic variability in retinal responses even under fixed stimuli. We contrast this with a classical null hypothesis of deterministic encoding and propose experiments to further test our uncertainty relation. By re-framing the retina as a probabilistic measurement device, we lay the foundation for future models of cortical dynamics rooted in quantum-like computation. We are not claiming that the brain could be working as a quantum-system, but rather putting forth the argument that the brain as a classical system could still implement quantum-inspired computations. We define quantum-inspired computation as a scheme that includes both probabilistic and time-sensitive computation, clearly separating it from classically implementable probabilistic systems.