Answers are grounded in this paper and cite their sources.
Search is a fundamental capability of intelligent systems, supporting information acquisition, candidate exploration, hypothesis discovery, solution generation, problem solving, and decision-making. Existing research has investigated various forms of individual search, including autonomous search by individual artificial intelligence models, collaborative search involving multiple agents, and human-directed information seeking. However, these approaches are frequently developed and evaluated through different architectural assumptions, and a general framework that integrates individual model search, multi-model collaborative search, and human search while dynamically determining which approach should be employed remains an important research direction. This paper proposes the Universal Self-Collaborative-Human Adaptive Search Framework (USCHASF), a general theoretical architecture for organizing search activities across artificial intelligence models, autonomous agents, human individuals, and heterogeneous human–artificial participant configurations. The framework establishes three fundamental search dimensions: Self-Search, in which an individual model or participant independently explores candidate answers, hypotheses, information sources, or solution spaces; Collaborative Search, in which multiple models or artificial agents distribute, coordinate, parallelize, exchange, and integrate search activities; and Human Search, in which human individuals or groups independently perform, direct, guide, expand, or participate in search processes. Its central contribution is the integration of Universal Adaptive Search Selection, Adaptive Search Mechanism Switching, Adaptive Search Combination, and Adaptive Search Reconfiguration. Rather than permanently relying on one search configuration, a system can select Self-Search, Collaborative Search, Human Search, or hybrid configurations according to task complexity, search progress, information uncertainty, evidence coverage, search quality, participant capabilities, resource availability, and operational constraints. It can also transition between search mechanisms during an ongoing search process, increase or reduce participant involvement, and reorganize search responsibilities as new requirements emerge. The framework further distinguishes search objectives, search spaces, search strategies, search participants, search mechanisms, and search configurations, enabling a mechanism-independent representation applicable to information retrieval, scientific discovery, knowledge exploration, hypothesis generation, web search, database exploration, reasoning, planning, and optimization. A formal model, adaptive decision architecture, switching policy, operational lifecycle, theoretical propositions, evaluation methodology, and governance principles are developed. To the best of our knowledge, this paper is the first to formally unify Self-Search, Collaborative Search, and Human Search within a general architecture that explicitly integrates adaptive search selection, dynamic switching, combination, and structural reconfiguration across heterogeneous human and artificial participants. The framework establishes a theoretical foundation for developing search systems that can dynamically determine not only what to search for, but also who should search, how search should be organized, and when the search mechanism should change.
No takes yet. Share an insight, caveat, or question.
Jincheng Zhang (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: