企业任务不该全靠大模型，应拆解为知识库+符号计算+模型提取的模块化架构。
Position: Avoid Overstretching LLMs for every Enterprise Task
- 将大模型视为接口，关键任务交由知识库与符号程序处理
- 理论证明有限容量模型无法覆盖企业知识广度，存在固有瓶颈
- 模块化设计更可靠、可维护，适合对稳定性要求高的企业场景
企业工作负载以确定性、结构化且依赖知识的任务为主，受成本、延迟和可靠性严格约束。尽管常通过大语言模型部署或蒸馏为小模型解决，但此类做法效率低、不可靠，且与任务结构不匹配。本文主张将语言模型视为接口而非整体引擎，将知识与计算外化至专用组件，以提升可靠性、可扩展性和透明性。理论分析表明，有限容量模型无法完整捕获企业任务所需的知识广度，导致效率与可解释性的内在限制。基于此，我们提出：大模型应仅用于确定性流程中的结构化信息提取，而计算与存储应交由知识库与符号程序承担。形式化证明显示，这种模块化架构比单一框架更具可靠性与可维护性，为企业的可持续任务处理提供坚实基础。
原文摘要 · Abstract (English)
Enterprise workloads are dominated by deterministic, structured, and knowledge-dependent tasks operating under strict cost, latency, and reliability constraints. While these are often addressed through large language model (LLM) deployment or distillation into smaller models, we argue this is inefficient, unreliable, and misaligned with enterprise task structures. Instead, AI systems should treat language models as interfaces rather than monolithic engines, externalizing knowledge and computation into dedicated components for greater reliability, scalability, and transparency. Our theoretical evidences show that finite-capacity models cannot fully capture the breadth of knowledge required for enterprise tasks, creating inherent limits to efficiency and interpretability. Building on this, we take the position that language models should primarily be used for structured extraction in deterministic enterprise workflows, while computation and storage are delegated to knowledge bases and symbolic procedures. We formally demonstrate that such modular architectures are more reliable and maintainable than monolithic frameworks, offering a sustainable foundation for enterprise tasks.
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。