Forecasting Scientific Progress with Artificial Intelligence

TL;DR

This paper introduces a new benchmark and evaluation framework to assess AI's ability to forecast scientific progress, revealing current models' limitations in predicting the timing and realization of scientific advances across disciplines.

Contribution

The paper presents CUSP, a comprehensive benchmark for scientific forecasting, and provides an empirical analysis of AI models' capabilities and shortcomings in predicting scientific progress.

Findings

AI models can identify plausible research directions but struggle to predict if and when advances will occur.

Performance varies across scientific domains, with better timing predictions in AI compared to biology, chemistry, and physics.

Models exhibit overconfidence and response biases, indicating unreliable uncertainty estimates.

Abstract

Artificial intelligence (AI) is increasingly embedded in scientific discovery, yet whether it can anticipate scientific progress remains unclear. To study this question, we introduce a temporally grounded evaluation framework for forecasting scientific progress under controlled knowledge constraints. We present CUSP (Cutoff-conditioned Unseen Scientific Progress), a multi-disciplinary and event-level benchmark that evaluates scientific forecasting in AI systems through feasibility assessment, mechanistic reasoning, generative solution design, and temporal prediction. Across 4,760 scientific events, we observe systematic and domain-dependent limitations in current frontier models. While models can identify plausible research directions from competing candidates, they fail to reliably predict whether scientific advances will be realized and systematically misestimate when they will occur. Performance is highly heterogeneous across domains, with the timing of AI progress more predictable than advances in biology, chemistry, and physics. Performance is largely insensitive to whether events occur before or after the training cutoff, suggesting these limitations cannot be explained solely by knowledge exposure in training data. Under controlled information access, additional pre-cutoff knowledge improves performance but does not close the gap to full-information settings, which becomes more pronounced for high-citation advances. Models also exhibit systematic overconfidence and strong response biases, indicating unreliable uncertainty estimation. Taken together, current AI systems fall short as predictive tools for scientific progress. Access to prior knowledge does not translate into reliable forecasting, and performance benefits more from post-event information than from forward-looking prediction.