OmniScientist: Toward a Co-evolving Ecosystem of Human and AI Scientists
Chenyang Shao, Dehao Huang, Yu Li, Keyu Zhao, Weiquan Lin, Yining Zhang, Qingbin Zeng, Zhiyu Chen, Tianxing Li, Yifei Huang, Taozhong Wu, Xinyang Liu, Ruotong Zhao, Mengsheng Zhao, Xuhua Zhang, Yue Wang, Yuanyi Zhen, Fengli Xu, Yong Li, Tie-Yan Liu
2025-11-24
Summary
This paper introduces OmniScientist, a new AI system designed to act as an 'AI Scientist' that goes beyond simply performing scientific tasks. It aims to recreate the *way* science is actually done, which involves collaboration, building on existing knowledge, and getting feedback from others.
What's the problem?
Current AI systems that try to do science treat it like a puzzle to be solved in isolation. They don't account for the fact that real scientific progress happens through teamwork, sharing ideas, and a system of checking each other's work (like peer review). Because they miss these key parts of how science works, these AI systems can't really integrate into the scientific community or create a lasting cycle of innovation.
What's the solution?
The researchers built OmniScientist, which isn't just about *doing* science, but also about *organizing* science. It includes a way to organize scientific knowledge based on how papers cite each other, a system for AI agents (and even human scientists) to work together on projects, and a platform called ScienceArena where research can be evaluated by others in a fair, anonymous way. This setup allows the AI to learn from existing research, collaborate effectively, and improve over time through feedback.
Why it matters?
This work is important because it moves AI scientists closer to being true partners in scientific discovery. By modeling the social and collaborative aspects of science, OmniScientist has the potential to accelerate research, create a more sustainable innovation process, and allow AI to genuinely contribute to and benefit from the human scientific community.
Abstract
With the rapid development of Large Language Models (LLMs), AI agents have demonstrated increasing proficiency in scientific tasks, ranging from hypothesis generation and experimental design to manuscript writing. Such agent systems are commonly referred to as "AI Scientists." However, existing AI Scientists predominantly formulate scientific discovery as a standalone search or optimization problem, overlooking the fact that scientific research is inherently a social and collaborative endeavor. Real-world science relies on a complex scientific infrastructure composed of collaborative mechanisms, contribution attribution, peer review, and structured scientific knowledge networks. Due to the lack of modeling for these critical dimensions, current systems struggle to establish a genuine research ecosystem or interact deeply with the human scientific community. To bridge this gap, we introduce OmniScientist, a framework that explicitly encodes the underlying mechanisms of human research into the AI scientific workflow. OmniScientist not only achieves end-to-end automation across data foundation, literature review, research ideation, experiment automation, scientific writing, and peer review, but also provides comprehensive infrastructural support by simulating the human scientific system, comprising: (1) a structured knowledge system built upon citation networks and conceptual correlations; (2) a collaborative research protocol (OSP), which enables seamless multi-agent collaboration and human researcher participation; and (3) an open evaluation platform (ScienceArena) based on blind pairwise user voting and Elo rankings. This infrastructure empowers agents to not only comprehend and leverage human knowledge systems but also to collaborate and co-evolve, fostering a sustainable and scalable innovation ecosystem.