Answers are grounded in this paper and cite their sources.
Experience sharing is a fundamental mechanism through which individuals and intelligent systems communicate knowledge acquired from previous activities, observations, decisions, successes, failures, and problem-solving processes. Although experience sharing has been investigated in multi-agent reinforcement learning, organizational knowledge management, collaborative learning, human-agent interaction, and artificial intelligence memory systems, existing approaches generally address particular application domains, participant configurations, or information-transfer mechanisms. This paper proposes Universal Multi-Participant Experience Sharing (UMPES), a unified conceptual framework designed to generalize experience sharing across arbitrary combinations of human participants, artificial intelligence agents, and hybrid human–AI entities. Unlike approaches restricted to agent-to-agent experience exchange or human-agent memory collaboration, UMPES conceptualizes experience sharing as an independent, configurable interaction capability that accommodates arbitrary participant cardinalities, communication directions, temporal relationships, experience categories, and sharing configurations. The framework encompasses unidirectional, bidirectional, multidirectional, individual-to-group, group-to-individual, group-to-group, cross-user, cross-agent, human-to-agent, agent-to-human, and temporally self-directed experience sharing. It further introduces a structured conceptual architecture covering experience acquisition, representation, selection, transmission, reception, interpretation, assessment, integration, retention, application, feedback, and redistribution. Particular attention is given to experiential provenance, contextual applicability, reliability, conflicting experiences, privacy, authorization, and the distinction between sharing an experience and successfully learning from it. This paper positions UMPES as a general theoretical foundation for experience-sharing interactions across heterogeneous participants rather than as a task-specific learning algorithm. To the best of our knowledge, this work proposes the first unified conceptual framework explicitly integrating these participant-independent, direction-independent, and cardinality-independent dimensions of experience sharing into a single human–artificial intelligence interaction model.
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: