Toward an Engineering of Science: Rebalancing Generation and Verification in the Age of AI

TL;DR

This paper addresses the challenge of unreliable AI-generated scientific artifacts by proposing structured 'blueprints' to improve verification efficiency and restore the balance between generation and verification costs in scientific epistemic infrastructure.

Contribution

It introduces 'blueprints', a structured research artifact format, as a novel engineering solution to enhance verification and mitigate epistemic pollution in AI-assisted science.

Findings

Blueprints enable more efficient verification of scientific claims.

Prototype implementation demonstrates feasibility of structured artifacts.

Rebalancing generation and verification costs can reduce epistemic pollution.

Abstract

AI systems can now cheaply generate plausible scientific artifacts such as papers, reviews, and surveys. This creates a risk of \emph{epistemic pollution} in our scientific systems, where unreliable but plausible-looking artifacts can accumulate faster than the system can filter them out. The problem is structural: the epistemic infrastructure of science was calibrated to a world where producing a plausible artifact required substantial expertise, labor, and time, so generation cost itself served as a rough filter; AI weakens that filter without comparably lowering verification cost. We argue that \textbf{AI-era science should treat this as an engineering problem: redesigning epistemic infrastructure to rebalance the costs of generation and verification}. The current paper-centered system makes verification expensive: papers compress long-context scientific logic into prose, forcing reviewers, human or AI, to reconstruct underlying argument structure before they can evaluate it. As one step in this direction, we propose \textbf{blueprints} as preliminary epistemic infrastructure: structured, decomposed research artifacts that represent claims, evidence, assumptions, and definitions as typed graph components. Blueprints are designed to trade an upfront generation cost for cheaper, more local, more distributed verification downstream. We have instantiated the proposal in a proof-of-concept prototype.