应对可持续AI三元悖论：基于大语言模型智能体与RAG的案例研究
Addressing the sustainable AI trilemma: a case study on LLM agents and RAG
AI总结：
该研究提出可持续AI三元悖论概念，通过LLM智能体与RAG的案例分析内存模块的能耗成本，提出量化能耗与性能权衡的新指标，揭示现有内存增强框架的能源低效问题，为可持续AI系统开发提供参考。
摘要：
大语言模型（LLM）已展现出强大能力，但其广泛部署与更高级应用带来了严峻的可持续性挑战，尤其是推理能耗问题。我们提出“可持续AI三元悖论”概念，揭示AI能力、数字公平与环境可持续性三者间的张力关系。通过针对LLM智能体与检索增强生成（RAG）的系统性案例研究，我们分析了内存模块设计中隐含的能源成本，并提出了全新指标以量化能耗与系统性能之间的权衡。实验结果显示，当前内存增强框架存在显著的能源低效问题，且资源受限环境面临的效率损失更为严重。本研究的发现对当前智能体设计中以LLM为中心的主流范式提出了挑战，也为开发更具可持续性的AI系统提供了实践洞见。
英文摘要：
Large language models (LLMs) have demonstrated significant capabilities, but their widespread deployment and more advanced applications raise critical sustainability challenges, particularly in inference energy consumption. We propose the concept of the Sustainable AI Trilemma, highlighting the tensions between AI capability, digital equity, and environmental sustainability. Through a systematic case study of LLM agents and retrieval-augmented generation (RAG), we analyze the energy costs embedded in memory module designs and introduce novel metrics to quantify the trade-offs between energy consumption and system performance. Our experimental results reveal significant energy inefficiencies in current memory-augmented frameworks and demonstrate that resource-constrained environments face disproportionate efficiency penalties. Our findings challenge the prevailing LLM-centric paradigm in agent design and provide practical insights for developing more sustainable AI systems.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
