對智能體模型的批判

摘要

什麼是代理？什麼構成代理性？隨著大型語言模型（LLM）系統被包裝成「編碼代理」、「AI共同科學家」等「具備代理性」的工具，宣稱能提升生產力，同時也引發了諸多「存在性」擔憂，例如AI在猜測中的「機器代理性」下可能脫離人類控制、造成毀滅性後果——此時釐清自動化與代理性的界線，無論是為了建構能力更強的系統，還是為了理解我們是否該恐懼、恐懼什麼，都變得至關重要。本文借鑒笛卡兒將代理性奠基於獨立思考的觀點，以及科幻作品中對自主存在的描繪，綜述當前AI代理的發展現狀，並從五個維度分析代理架構：目標、身份、決策、自我調節與學習。具體而言，我們主張真正的代理性需要這些結構內化於系統本身，而非透過外部支架拼湊而成。此區別界定了兩類系統：一類是「代理系統」，其能力源於工程化工作流程；另一類是「具代理性系統」，其能力（包括社交互動）是內生湧現的。前者專為預設任務設計，後者則能在開放世界中具備真正的自主性。基於此分析，我們提出通用代理模型的目標-身份-配置器（GIC）架構，結合分層目標分解、身份演化、基於獨立訓練的世界模型進行模擬推理、學得的自我調節，以及從真實與模擬經驗中進行的自我導向學習。此外，我們就具備更高自主性與「代理性」但仍受人類監督的系統，分享其在可審計性、可控性與安全性方面的見解。

English

What is an agent? What constitutes agency? With the rise of Large Language Model (LLM) systems marketed as ``coding agents'', ``AI co-scientists'', and other ``agentic" tools that promise to drive up productivity, and at the same time, ``existential" concerns such as AI escaping human control with destructive power under a speculative ``machine agency" against humans, it has become essential to clarify where automation ends and agency begins, both for building capable systems and for understanding whether and what to fear. Drawing on Descartes' grounding of agency in independent thought, and on portrayals of autonomous beings in science fiction, we survey the current landscape of AI agents, and analyze agent architectures along five dimensions: goal, identity, decision-making, self-regulation, and learning. Specifically, we argue that genuine agency requires these structures to be internalized within the system itself rather than assembled through external scaffolding. This distinction between agentic systems, whose competence resides in engineered workflows, and agentive systems, whose capabilities (including social interaction) arise endogenously, defines the boundary between systems designed for prescribed tasks, and those capable of operating in the open world with true autonomy. Building on this analysis, we propose the Goal-Identity-Configurator (GIC) architecture for a general-purpose agent model, combining hierarchical goal decomposition, identity evolution, simulative reasoning grounded in a separately trained world model, learned self-regulation, and self-directed learning from both real and simulated experience. Furthermore, we share insight on the auditability, controllability, and safety of agentive systems that possess greater autonomy and ``agency", but remain under human oversight.