Large language models have been demonstrated to be valuable in different fields. ChatGPT, developed by OpenAI, has been trained using massive amounts of data and simulates human conversation by comprehending context and generating appropriate responses. It has garnered significant attention due to its ability to effectively answer a broad range of human inquiries, with fluent and comprehensive answers surpassing prior public chatbots in both security and usefulness. However, a comprehensive analysis of ChatGPT's failures is lacking, which is the focus of this study. Ten categories of failures, including reasoning, factual errors, math, coding, and bias, are presented and discussed. The risks, limitations, and societal implications of ChatGPT are also highlighted. The goal of this study is to assist researchers and developers in enhancing future language models and chatbots.

翻译：大型语言模型已被证明在不同领域具有重要价值。OpenAI开发的ChatGPT通过海量数据训练，能理解上下文并生成适当回应，模拟人类对话。由于其能够以流畅且全面的回答有效处理广泛的人类提问，在安全性和实用性上均超越此前公开的聊天机器人，因而备受关注。然而，针对ChatGPT失败案例的系统性分析尚属空白——这正是本研究的核心。我们提出并讨论了十类失败案例，包括推理错误、事实性错误、数学计算、编程缺陷及偏见等。同时强调了ChatGPT的风险、局限性及其社会影响。本研究旨在帮助研究人员和开发者改进未来语言模型与聊天机器人。