Overnight
Between midnight and 6am, Googlebot crawls 21,000 pages, Bingbot crawls 10,000, and the comprehensive check grinds through 15,000. The site is more active at 3am than at 3pm.
AI & Technology寫作
關於設計、開發、AI基礎設施和產品構建的思考。
When Your Agent Becomes the Researcher
Claude Code discovered adversarial attacks that outperform 30+ existing methods. Autoresearch just crossed from experiment to capability.
AI & TechnologyCompound Context: Why AI Projects Get Better the Longer You Stay With Them
Every problem you solve with an AI agent deposits context that the next session withdraws with interest. This is context compounding.
AI & TechnologyThe Supply Chain Is the Attack Surface
Trivy got compromised. Then LiteLLM. Then 47,000 installs in 46 minutes. The AI supply chain worked exactly as designed.
AI & TechnologyAI Agent Memory Architecture That Actually Works
Hybrid BM25+vector retrieval, skills as markdown, drift detection. Five March 2026 papers validate the same architecture built from production failures.
AI & TechnologyDeploy and Defend: The Agent Trust Paradox
Every major institution is racing to deploy AI agents and building walls against them simultaneously. 1 in 8 enterprise AI breaches now involve agents.
AI & TechnologyMidjourney V8 Killed Your V7 Workflow: What Actually Changed
V8 isn't a better V7. It's a different creative loop: personalization over prompting, native 2K, conversation mode, and a cost model that punishes the old workflow.
AI & TechnologyEvery Iteration Makes Your Code Less Secure
43.7% of LLM iteration chains introduce more vulnerabilities than baseline. Adding SAST scanners makes it worse. SCAFFOLD-CEGIS cuts degradation to 2.1%.
AI & Technology您的代理沙箱只是建議
攻擊者開了一個GitHub議題，並在Cline的下一個版本中植入惡意軟體。代理沙箱在三個層級上失敗。以下是真正有效的防禦方式。
AI & Technology隱形代理：為何您無法治理看不見的事物
Anthropic悄悄在使用者的Mac上放置了10GB的虛擬機。代理可觀測性需要三個層次：資源計量、策略執行與執行時期稽核。
AI & Technology靜默外洩：您未曾建構的攻擊面
一個惡意網頁將指令注入URL元資料中。代理程式擷取了該頁面、讀取了毒化內容，並將API金鑰外洩。沒有錯誤。沒有日誌。
AI & Technology工作階段就是提交訊息
Git 記錄了什麼改變了。Agent 工作階段記錄了為什麼改變。當 agent 撰寫程式碼時，工作階段記錄才是真正的設計文件——而我們卻將它丟棄了。
AI & Technology外部出版物
技術寫作： Introl
Local LLM Hardware Pricing Guide 2025
本地運行大型語言模型的全面硬體建議和成本分析。
H100 vs. H200 vs. B200: Choosing the Right NVIDIA GPUs for Your AI Workload
針對不同AI工作負載比較NVIDIA最新資料中心加速器的GPU選擇指南。
TPU Architecture: Complete Guide to Google's 7 Generations
深入技術分析Google張量處理單元從TPUv1到TPUv5的演進。
GPU Memory Pooling in Multi-Tenant Kubernetes
容器化環境中GPU叢集的資源共享策略。
Ray Clusters: Distributed AI Computing Infrastructure
使用Ray框架構建和管理分散式AI運算的指南。
DeepSeek V3 Open Source AI Cost Advantage
開源LLM經濟學分析和DeepSeek的競爭定位。
NVIDIA Vera Rubin GPU 600kW Racks 2027
未來資料中心電力需求和NVIDIA下一代GPU路線圖。
SMR Nuclear Power for AI Data Centers
為下一代AI基礎設施供電的小型模組化反應爐解決方案。
DeepSeek MHC Architecture Breakthrough
DeepSeek多頭壓縮架構創新的技術分析。