Position: AI for Science Should Treat Measurement-to-Dataset Pipelines as Inference Components

AI for Science (AI4Science) workflows often treat the released dataset as a fixed interface to the underlying system. However, in domains relying on \emph{indirect observation}, the learner observes a derivative representation produced by multi-stage measurement, reconstruction, and preprocessing pipelines. $\mathbf{\text{We}}\mathbf{\text{argue}}\mathbf{\text{that}}\mathbf{\text{these}}\mathbf{\text{measurement-to-dataset}}\mathbf{\text{pipelines}}\mathbf{\text{are}}\mathbf{\text{inference}}\mathbf{\text{components:}}\mathbf{\text{treating}}\mathbf{\text{their}}\mathbf{\text{outputs}}\mathbf{\text{as}}\mathbf{\text{“given}}\mathbf{\text{data”}}\mathbf{\text{freezes}}\mathbf{\text{an}}\mathbf{\text{observation}}\mathbf{\text{model}}\mathbf{\text{and}}\mathbf{\text{obscures}}\mathbf{\text{uncertainty}}\mathbf{\text{over}}\mathbf{\text{feasible}}\mathbf{\text{pipeline}}\mathbf{\text{choices.}}$ We identify three failure modes arising from this ``frozen lens’': $\mathbf{\text{(C1)}}\mathbf{\text{hidden}}\mathbf{\text{hypothesis}}\mathbf{\text{space}}$, where the released dataset does not specify the pipeline configuration or its validity conditions; $\mathbf{\text{(C2)}}\mathbf{\text{uncertified}}\mathbf{\text{transportability}}$, where a pipeline may be documented but its regime of validity is untested, so failures under distribution shift cannot be adjudicated; $\mathbf{\text{(C3)}}\mathbf{\text{ungoverned}}\mathbf{\text{multiplicity}}$, where many defensible pipelines exist and dispersion is real but not propagated into uncertainty-aware evidence. We stress-test these claims with a large-scale neuroscience empirical audit, finding a survival rate of $\approx 0.0004\mathrm{\%}$ under a cross-dataset stability criterion. We call on the AI4Science community to make pipelines \emph{computable} inference objects via domain-specific Computable Observation Frameworks. This shift enables quantifying pipeline adequacy and stability, converting implicit implementation choices into auditable, reproducible, and cumulative scientific evidence.