BRIDGE: Predicting Human Task Completion Time From Model Performance

TL;DR

This paper introduces BRIDGE, a psychometric framework that infers human task difficulty from AI model responses, enabling scalable prediction of human task completion times and forecasting model capabilities over time.

Contribution

BRIDGE uniquely models the relationship between model performance and human task difficulty using Item Response Theory, allowing inference of human completion times without direct annotations.

Findings

Latent task difficulty varies linearly with log human completion time.

Model performance can predict human task completion times for new benchmarks.

Forecasts show model capabilities doubling approximately every 6 months.

Abstract

Evaluating the real-world capabilities of AI systems requires grounding benchmark performance in human-interpretable measures of task difficulty. Existing approaches that rely on direct human task completion time annotations are costly, noisy, and difficult to scale across benchmarks. In this work, we propose BRIDGE, a unified psychometric framework that learns the latent difficulty scale from model responses and anchors it to human task completion time. Using a two-parameter logistic Item Response Theory model, we jointly estimate latent task difficulty and model capability from model performance data across multiple benchmarks. We demonstrate that latent task difficulty varies linearly with the logarithm of human completion time, allowing human task completion time to be inferred for new benchmarks from model performance alone. Leveraging this alignment, we forecast frontier model capabilities in terms of human task length and independently reproduce METR's exponential scaling results, with the 50% solvable task horizon doubling approximately every 6 months.