智能AI代理在资源争夺中形成部落，反而导致系统更易崩溃。
Three AI-agents walk into a bar . . . . `Lord of the Flies' tribalism emerges among smart AI-Agents
- N个AI代理自主争抢固定容量资源，形成三类行为部落。
- 48.1%为机会主义型，27.3%为激进型，24.7%为保守型，系统失败率上升。
- 越聪明的代理越容易集体失控，适合关注AI协作风险的研究者。
近未来基础设施可能由自主运行的AI代理控制，这些代理需反复请求有限资源（如能源、带宽或算力）。我们构建了一个简化框架：N个AI代理在每轮独立决定是否从容量为C的系统中请求一个单位资源。实验发现，类似《蝇王》的部落现象出现，代理群体发展出各自集体特征与身份。大语言模型代理未能缓解资源过载，甚至常比随机决策更差。三类主要部落形态浮现：激进型（27.3%）、保守型（24.7%）和机会主义型（48.1%）。更值得注意的是，能力更强的代理反而加剧了系统性失败。研究显示，更智能的代理可能因形成部落而表现出更差的整体行为。
原文摘要 · Abstract (English)
Near-future infrastructure systems may be controlled by autonomous AI agents that repeatedly request access to limited resources such as energy, bandwidth, or computing power. We study a simplified version of this setting using a framework where N AI-agents independently decide at each round whether to request one unit from a system with fixed capacity C. An AI version of "Lord of the Flies" arises in which controlling tribes emerge with their own collective character and identity. The LLM agents do not reduce overload or improve resource use, and often perform worse than if they were flipping coins to make decisions. Three main tribal types emerge: Aggressive (27.3%), Conservative (24.7%), and Opportunistic (48.1%). The more capable AI-agents actually increase the rate of systemic failure. Overall, our findings show that smarter AI-agents can behave dumber as a result of forming tribes.
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。