An engineering notebook on AI, systems, and everyday experiments.

记录在API因合规问题被拒后，使用视觉大模型从截图提取财务数据，实现十年手动记账流程的自动化。展示了本地模型、交叉验证和人机协作工作流如何安全处理敏感金融数据。

Documents automating a decade-long manual financial reconciliation process using vision LLMs when API access was blocked by compliance. Demonstrates using local models with screenshots, cross-validation, and human-in-the-loop workflows to process sensitive financial data.

GPT-5是产品升级而非单纯模型升级：新增reasoning_effort和verbosity参数，可控性大幅提升，让开发者能根据场景灵活权衡推理深度和回复长度。

GPT-5 is a major product upgrade, not just a model upgrade: new API parameters for reasoning_effort and verbosity enable unprecedented controllability for building AI-powered products.

记录通过迭代式问题解决开发自动截图功能的经历。关键经验：优化问题定义而非仅优化提示词、从"最小可行真相"出发锚定工作流、根据任务需求匹配合适的AI模型能力边界。

Recounts building an automated screenshot feature through iterative problem-solving with different AI models. Key lessons: refine the problem not just the prompt, anchor in a "Minimum Viable Truth," and match AI capabilities to task requirements.

在真实编程和调研任务中评测Kimi K2的Agentic能力：执行韧性出色，适合作为信息采集前端；但工具调用稳定性和生态适配仍有提升空间。

Testing Kimi K2 as an agentic model in real coding and research tasks: excellent execution resilience makes it ideal for information gathering, but tool integration friction needs improvement for production use.

记录通过语音输入、全天候录音、可穿戴相机和Agentic系统拓展与AI沟通带宽的实验。提出"赛博长生"作为务实的生命观，通过提升生命密度、质量和边界来放大创造力与影响力。

Chronicles experiments with voice input, all-day recording, wearable cameras, and agentic systems to expand AI communication bandwidth. Introduces "Cybernetic Immortality" as a pragmatic philosophy to increase life density, quality, and cognitive boundaries.