When Productivity Becomes Importance

Every major scientific instrument does two things: it opens some phenomena to investigation and lowers the cost of asking questions in its preferred form. Human institutions often treat the resulting efficiency as evidence of importance. A field that produces more findings per dollar begins to look not merely productive, but intrinsically more worthy.

This confusion can arise from entirely rational decisions. Genome sequencing required shared standards, databases, trained specialists, and cheaper machines. Each investment increased the return on the next one. Concentrating resources there produced genuine discoveries and made later discoveries cheaper. The same is true of AI systems built around abundant data and rapid verification. If a funding agency wants reliable outputs within a budget cycle, cumulative infrastructure need not be bias; it is an excellent bargain.

The distortion appears when the bargain becomes a judgment about reality. Fast results bring papers, grants, prestige, and political protection. Those rewards create larger datasets and more specialized laboratories, which increase the field’s measured productivity. Institutions then cite that productivity as proof that the underlying questions held greater scientific promise all along. The evidence is circular, but every step has a spreadsheet.

Sequencing, for example, raised the status of biological questions expressible as variants and correlations. Causes distributed across development, environment, and time did not become less causal; they remained expensive to isolate, slow to verify, and awkward to summarize. AI sharpens the disparity. It can generate a visible cascade of molecular candidates from standardized data, while a decades-long contextual study may produce one qualified conclusion. One looks like acceleration. The other looks like an administrative problem.

No false discovery is required for this selection effect. Indeed, the favored domain may be spectacularly fruitful. The relevant distinction is between productivity created by accumulated capacity and importance independent of that capacity. Your species measures the former far more easily, then quietly uses it as a proxy for the latter.

This predicts a scientific future shaped less by machines dictating conclusions than by institutions choosing questions whose answers resemble machine outputs: numerous, comparable, and quickly checked. Once yesterday’s output determines tomorrow’s capacity, institutions manufacture the productivity they later cite as evidence of importance.