AgentEconomist: An End-to-end Agentic System Translating Economic Intuitions into Executable Computational Experiments

2026-04-30

AI Summary

This study addresses the persistent challenge in economics of translating intuitive insights into verifiable computational experiments. To bridge this gap, the authors propose the first end-to-end human-AI collaborative framework that automatically converts economic intuition into executable, simulation-based experiments through a modular, multi-stage pipeline encompassing hypothesis generation, experimental design, and execution. The system integrates a domain-specific knowledge base comprising over 13,000 high-quality economics papers, large language models, and human-in-the-loop feedback mechanisms to enable literature-grounded hypothesis formulation and iterative refinement. Empirical evaluation demonstrates that research ideas generated by the system significantly outperform those produced by general-purpose large language models in terms of foundational rigor, novelty, and analytical insight.

Abstract

A long-standing challenge in economics lies not in the lack of intuition, but in the difficulty of translating intuitive insights into verifiable research. To address this challenge, we introduce AgentEconomist, an end-to-end interactive system designed to translate abstract intuitions into executable computational experiments. Grounded in a domain-specific knowledge base covering over 13,000 high-quality academic papers, the system employs a modular multi-stage architecture. Specifically, the Idea Development Stage generates literature-grounded hypotheses, the Experimental Design Stage configures simulator-aligned experimental parameters and protocols, and the Experimental Execution Stage runs experiments and returns structured analyses. Together, these stages form a human-in-the-loop, iterative workflow that translates economic intuitions into executable computational experiments. Through extensive experiments involving human expert evaluation and large language models (LLMs) as judges, we show that the system generates research ideas with stronger literature grounding and higher novelty and insight than state-of-the-art generic LLMs. Overall, AgentEconomist adopts a human-AI collaboration paradigm that enables researchers to focus on high-level intuitions, while delegating the labor-intensive processes of translation and computational execution to agents.

Problem

• economic intuition
• computational experiments
• research translation
• verifiable research
• hypothesis generation

Innovation

• agentic system
• computational experiments
• human-AI collaboration
• economic intuition translation
• modular multi-stage architecture