📄 中文摘要
每次与人工智能的对话最终都会重置，导致上下文的丧失。无论是Claude还是ChatGPT，都会在消息数量或上下文窗口达到极限后，回到空白提示。这种重置虽然令人烦恼，但更大的问题在于你失去了之前建立的上下文，包括讨论的决策、解释的研究和提供的背景信息。每次重置后，都需要重新解释这些内容，造成了效率的下降。尽管消息限制和上下文窗口是大型语言模型的固有特性，未来会有改进和解决方案，但上下文的丧失依然是一个亟待解决的问题。
你的人工智能达到了极限，但你的知识不应该
出处: Your AI Hit Its Limit. Your Knowledge Shouldn't.
发布: 2026年3月27日
📄 English Summary
Your AI Hit Its Limit. Your Knowledge Shouldn't.
Every AI conversation eventually resets, leading to a loss of context. Whether using Claude or ChatGPT, users will find themselves back at a blank prompt after reaching message limits or context window capacities. This reset is mildly annoying, but the real issue lies in losing the context built up during the conversation, including decisions discussed, research explained, and background provided. Each reset necessitates re-explaining these details, resulting in decreased efficiency. While message limits and context windows are inherent characteristics of large language models, future improvements and workarounds may not fully address the persistent problem of context loss.