OmniScientist: Toward a Co-evolving Ecosystem of Human and AI Scientists
🤖 AI Summary
Current AI scientist systems model scientific research as isolated optimization problems, neglecting its inherently social and collaborative nature. They lack explicit modeling of contribution attribution, peer review, and knowledge networks, hindering integration into real-world scientific ecosystems. This paper introduces the first AI scientist ecosystem grounded in human research practices, establishing an end-to-end scientific workflow encompassing data curation, literature synthesis, automated experimentation, paper generation, and structured peer review. Key contributions include: (1) a structured scientific knowledge graph encoding conceptual and relational semantics; (2) the Open Scientific Protocol (OSP), a multi-agent research coordination framework enabling role-aware, reproducible collaboration; and (3) ScienceArena—an open, human-AI co-review platform integrating citation networks, concept-based semantic linking, blinded reviewer matching, and Elo-style dynamic performance ranking. Empirical evaluation demonstrates substantial improvements in AI scientists’ collaborative innovation capability under complex, realistic conditions and enhanced interoperability with human research communities.
Application Category
📝 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.
Problem
Research questions and friction points this paper is trying to address.
Existing AI Scientists overlook the collaborative nature of scientific research
Current systems lack modeling of scientific infrastructure and human interaction
OmniScientist bridges this gap by encoding human research mechanisms
Innovation
Methods, ideas, or system contributions that make the work stand out.
Simulates human scientific system with structured knowledge
Enables multi-agent collaboration via research protocol
Implements open evaluation platform with user voting
🔎 Similar Papers
No similar papers found.
