A cell detects the direction of a chemical signal by counting receptor bindings. Molecules arrive from a distant source, diffuse through the medium, and land on receptors distributed across the cell surface. More molecules arrive from the direction facing the source. The cell reads the spatial pattern of arrivals to estimate where the signal is coming from.
Sun et al. (arXiv:2603.10207) show that the first few bindings carry more directional information than all subsequent bindings combined. The initial arrivals land on the receptors closest to the source, before diffusion has time to scatter molecules around the cell. As more molecules arrive, diffusion randomizes their landing pattern. The spatial signal degrades with accumulation. The cell's best estimate of source direction comes not from averaging many events but from weighting the earliest ones.
This is not a statistical effect that washes out with more data. It's structural. The first binding events occur in a state of maximum directional bias — the molecules haven't had time to explore the full receptor surface. Later events carry less information per event because diffusion has already mixed the directional signal with noise. Waiting longer doesn't help. More data is literally worse data.
An unrelated system displays the same structure. Macatangay and Strachan (arXiv:2603.10939) study the thermal decomposition of TATB, a high-energy crystal. The reaction proceeds through multiple steps: first a hydrogen transfer, then ring opening, then fragmentation. Nuclear quantum effects — the tunneling and zero-point energy that distinguish quantum particles from classical billiard balls — lower the activation barrier of the first step by approximately 8%. The same quantum effects have no measurable impact on later steps.
The explanation maps onto the cell-sensing case. The first reaction step occurs from the undisturbed crystal lattice, where quantum coherence is intact and the proton's wave function samples the full width of the potential barrier. After the first bond breaks, the local environment reorganizes. The lattice distorts. The potential energy surface becomes rough, asymmetric, dominated by classical thermal activation. Quantum effects that were decisive for the intact structure become negligible in the disrupted one.
In both cases, the first event in a sequence is structurally distinct from subsequent events of the same kind. Not because first events are inherently special, but because the first event occurs in the original state — the unperturbed state — and each subsequent event occurs in a state already modified by what came before. The cell surface before any binding is a clean sensor. After the first binding, receptor saturation begins. The crystal before any bond-breaking is quantum-mechanically coherent. After the first break, classical disorder dominates.
The pattern: the first event has access to structure that it then destroys by occurring. The information in the first receptor binding comes from the uncontaminated directional gradient, which the binding itself begins to obscure. The quantum advantage in the first reaction step comes from the intact lattice coherence, which the reaction itself disrupts. The event that extracts the most from the initial conditions is the event that eliminates those conditions.
This isn't primacy bias or anchoring. Those are cognitive effects where first impressions persist despite better evidence. This is physical: the initial state genuinely contains more accessible structure than any subsequent state, because the process of accessing it is also the process of degrading it. The measurement consumes what it measures — not through observer effects or uncertainty principles, but through the ordinary mechanics of the system responding to its own transitions.