AI Research Agents Narrow Scientific Exploration

arXiv:2605.27905 · 2026-05-28 공개 · arXiv · PDF

llm idea-generation scientific-exploration technical-methods research-areas ai-assisted-discovery recombination citation-impact

Abstract

AI research agents can now generate research ideas, design experiments, run code, and draft papers, raising the possibility of large-scale AI-assisted scientific discovery. Many current agent frameworks explicitly encourage the generation of novel and high-impact ideas. Yet it remains unclear whether AI-assisted ideation broadens scientific exploration or mainly concentrates around existing work. We study AI research agents as scientific search systems. Using four AI research-agent frameworks and six large language models, we generate 37,802 scientific ideas from shared seed literature across citation-defined research areas in AI and machine learning. We then compare the resulting AI ideas against human-authored papers from the same research areas, follow-on human research emerging from the same seed literature, and the seed literature itself. Across experiments, four consistent patterns emerge. First, AI-generated ideas are substantially more concentrated than human-authored papers from the same research areas. Second, AI-generated ideas remain much closer to their starting literature than later human follow-on work does. Third, papers most similar to AI-generated ideas tend to receive lower subsequent citations. Fourth, when AI-generated ideas differ from prior work, the differences arise primarily from recombining existing technical methods rather than introducing fundamentally new research questions. Overall, current AI research agents appear better suited to local elaboration than to broadening scientific exploration.

한국어 요약

📋 한 줄 요약

**[AI 연구 에이전트 / 과학 탐색]** AI 에이전트 4 프레임워크·6 LLM에서 37,802 아이디어 분석, AI 생성 아이디어가 인간 논문보다 substantially 집중·seed에 가까이 머물고 follow-on citation이 낮음, novelty의 대부분이 새 질문이 아닌 기존 method recombination.

🎯 핵심 기여도

💡 핵심 아이디어

AI 연구 에이전트는 ideation을 자동화하지만 그 출력은 통계적으로 집중·국소 elaboration에 치우치며, 새 연구 질문을 던지기보다 기존 method를 재조합하는 경향이 강해 현재 시스템은 과학 탐색을 broaden하기보다 좁히는 방향으로 작동한다.

🔬 기술적 접근법

📊 주요 결과

💭 의의 및 한계

**의의**: AI 연구 에이전트의 ideation 효과에 대한 첫 대규모 실증 진단, 4 프레임워크·6 LLM의 generality 확보, "broaden vs concentrate" 질문의 정량 답변, novelty의 origin 분해. **한계**: AI·ML 분야 중심 — 다른 과학 분야 일반화 미검증, 현재 세대 에이전트 분석으로 빠른 발전 가능성, citation을 quality proxy로 한 가정의 한계.

🚀 실용적 활용