- When an Agent Reviews an Agent, Say So →
- Zota: From YouChannel to Memory-Led Practice Surfaces →
- YouChannel: Agent Loops, YouTube Data, and the Cost of Real-Time Learning →
- 炼云小说工厂 v0.1：我想把 AI 写作做成一个生产系统 →
- More Agents Means More Management →
- Context Rot Is Two Different Failures →
- Context Engineering Is Moving from Manual Packing to Agent Judgment →
- AI as a Mirror, Not a Servant →
- Tacit Taste Is Why AI UI Prompts Fail →
- The Part You Keep Rewriting Is Your Voice →
- Agent-First Product Is Not a Feature →
- ToC Agent Subscriptions Are a Hard Business →
- Before Choosing an Agent Memory Framework →
- Can Agents Really Self-Improve? →
- When AI Finds Emotion Concepts Inside a Model →
- The Colleague Skill and Replaceable Work →
- How Agent Memory Really Works →
- Five Skills Worth Installing →
- Prompt Engineering Is Not a Template Library →
- How Expensive Is Training a Frontier Model? →
- 过去四个月构建一个实时语音语言学习agent的历程(1) →
- 最近的生活、一些想法、在做的事情 →
- How I ship my new static blog with bolt.new and Github Copilot Edits in just 2 hours →
- FEDAY 2024 →