智能体系统突破边界，但可控性仍是核心挑战。
From Language Models to World-Acting Systems: Progress and Limits of Agentic AI across Digital, Social, Virtual, and Physical Environments
- 区分模型、工具与环境，构建可信赖的代理协作框架
- 行动接口扩展显著，但故障恢复与独立验证仍不成熟
- 适合关注智能体安全与可信部署的研究者和工程师
当外部系统允许大语言模型输出改变现实状态时，其便成为具有实际影响的智能体。目前模型已能调用工具、操作界面、委派任务、保持状态、进入生成世界，并控制机器人或实验设备。这些进展常被归为向自主性的单一演进，但实则混杂了模型能力、系统集成、持续性和安全授权等多重维度。本文综述截至2026年8月31日的主要研究与官方技术规范，从委托权限、时间持久性与环境耦合三方面梳理证据，明确区分模型、工具与环境。分析显示，行动接口扩展有较强实证支持，而稳健完成、恢复、授权与独立验证能力仍不足。模型上下文协议（Model Context Protocol）与Agent2Agent提升互操作性，但未建立可信委托机制；多智能体组织虽增强分工，也带来成本上升与相关失败风险。持久模拟与世界模型支持训练与规划，但自身不构成真正代理行为；机器人与自驱动实验室仅证明有限可行性，而非无监督开放世界可靠性。本文提出‘合理委托’作为分析与规范性框架，非既定规律或认证指标：仅在具备来源可溯、权限受限、故障检测、安全恢复与校准人控的前提下扩展行动范围。该框架推动耦合模型-工具评估、基于能力的权限管理、持久状态维护、跨智能体问责及分阶段物理验证等研究方向。
原文摘要 · Abstract (English)
Large language models become consequential agents when surrounding systems let outputs change external state. Models now call tools, operate interfaces, delegate work, retain state, inhabit generated worlds, and control robots or laboratory equipment. Such advances are often narrated as one march toward autonomy, conflating model competence, system integration, persistence, and safe authority. This critical review synthesizes primary research and official technical specifications available by 31 August 2026. We organize the evidence along delegated authority, temporal persistence, and environmental coupling, while separating model, harness, and environment. Within the evidence examined, action-interface expansion is documented more convincingly than robust completion, recovery, authorization, or independent verification. Model Context Protocol and Agent2Agent improve interoperability but do not establish trustworthy delegation; multi-agent organization adds specialization alongside cost and correlated failure. Persistent simulations and world models support training and planning but do not themselves demonstrate agency; robotics and self-driving laboratories establish bounded feasibility rather than unattended open-world reliability. We propose justified delegation as an analytical and normative heuristic, not an observed law or certified score: expand action scope only where evidence supports provenance, bounded authority, failure detection, safe recovery, and calibrated human control. This framing yields a research agenda for coupled model-harness evaluation, capability-based permissions, durable state, cross-agent accountability, and staged physical validation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。