Large Language Models (LLMs) bring transformative benefits alongside unique challenges, including intellectual property (IP) and ethical concerns. This position paper explores a novel angle to mitigate these risks, drawing parallels between LLMs and established web systems. We identify "citation" as a crucial yet missing component in LLMs, which could enhance content transparency and verifiability while addressing IP and ethical dilemmas. We further propose that a comprehensive citation mechanism for LLMs should account for both non-parametric and parametric content. Despite the complexity of implementing such a citation mechanism, along with the inherent potential pitfalls, we advocate for its development. Building on this foundation, we outline several research problems in this area, aiming to guide future explorations towards building more responsible and accountable LLMs.

翻译：大型语言模型（LLMs）在带来变革性优势的同时，也伴随着独特的挑战，包括知识产权（IP）和伦理问题。本文从新颖视角探讨如何缓解这些风险，将LLMs与成熟的网络系统进行类比。我们发现"引用"是LLMs中缺失的关键要素，它能够提升内容透明度与可验证性，同时应对知识产权和伦理困境。我们进一步提出，面向LLMs的全面引用机制应同时涵盖非参数化和参数化内容。尽管实施此类引用机制存在复杂性及固有潜在缺陷，我们仍倡导其开发。在此基础上，我们概述了该领域的若干研究问题，旨在为构建更负责任、更可问责的LLMs的未来探索提供指引。