Tags

news (21) github (21) AI (19) security (9) open-source (11) trust (1) industry (1) supply-chain (2) LLM (10) quantization (3) 编码代理 (1) agi (1) 开发工具 (6) AI代理 (7) 记忆引擎 (1) 自进化 (1) Token压缩 (1) safety (2) inference (1) regulation (2) 技能框架 (1) 记忆系统 (1) 本地优先 (1) 工程化 (5) 开源替代 (1) 上下文压缩 (1) semiconductor (2) autonomous driving (1) machine learning (1) digital humanities (1) speculative decoding (1) distillation attack (2) model transparency (1) Claude Code (2) 开源趋势 (5) blog (24) agent (11) AI safety (3) 隐私 (2) 标准化 (1) 工具链 (1) flink (2) models (1) TypeScript (1) infrastructure (2) compilers (1) cpp (1) on-device AI (1) hardware (2) 技能库 (1) 本地化 (1) 开源工具 (1) open-weight (1) AI policy (1) mathematics (1) vLLM (1) MCP (3) Rust (1) Kimi K3 (1) agent-memory (1) agent-skills (1) llm-inference (1) npm (1) AI weekly (1) 开源模型 (2) 大语言模型 (1) AI 算力 (1) Agent (1) documentation (2) AI周报 (1) OpenRouter (1) Anthropic (1) 模型层商品化 (1) 供应链安全 (2) 本地模型 (1) ai-agent (3) 开源 (5) qa (1) 安全 (2) 硬件 (1) claude-code (2) codex (1) self-hosted (1) rust (1) prompt (1) dev-tools (1) trend (1) phd (1) OpenAI (1) 数学 (1) 芯片 (1) AI Agent (5) 自托管 (3) 本地推理 (1) AI安全 (3) AI智能体 (1) 国产算力 (1) rss (1) llm (1) ai (2) subscription (1) 模型定价 (1) 芯片供应链 (1) windows (1) 周报 (1) 算力 (1) 大模型 (1) trending (1) open-weights (1) AI-agents (1) rag (2) confidence (1) api (2) architecture (2) workflow (1) project (1) learning (1)

Tag: ai

AI 产出文档必须过人工审计

AI 产出的文档只是草稿:先定审计维度,人工审计确认后修改,才成为成熟文档。慢一点、严一点,比堆一堆不可信文档强。