Protein Observer

A constructed measurement-protocol companion to The Limits of Falsifiability

This is not a molecular-dynamics simulation. The code explicitly makes each click reduce a scalar uncertainty variable and sometimes trigger lock-in. Any displayed backaction is therefore a premise of this toy protocol, not evidence that every biological measurement changes its target in the same way.

Hydrophobic

Hydrophilic

Measured

Challenge: Track the Residue

System State

Measurements

Fold Progress

Uncertainty100%

Clicking is an intervention in this code. It multiplies the display's uncertainty variable by 0.7 and may trigger a thresholded lock-in. The resulting trade-off is constructed rather than discovered.

Endpoint underdetermination: Several toy histories can end at a similar display state, so the endpoint alone does not identify the path. That is a limitation of this record, not proof of irreducible uncertainty about real protein energy flows.

The causal question: one run cannot reveal its own no-intervention counterfactual. Repeated randomized controls, passive channels, calibrated probes, and mechanistic models can nevertheless estimate backaction and discriminate rival explanations. Whether a real assay perturbs its target is protocol-specific.