AI 教程与实战
收录 AI 工具教程、提示词、安装部署、配置指南、开发实战与可复现步骤。
分类文章5603
当前页94 / 141
持续更新永久归档
分类文章
按发布时间从新到旧 · 每页 40 条
教程 / 实战普通
Building a Sub-Second Enterprise RAG Engine with PostgreSQL, pgvector, and the Gemini API
The AI engineering space is currently obsessed with unbundling the data stack. The moment...
信息来源:DEV Community · Gemini
教程 / 实战普通
Repository Agent-Security Gap Study — v2h baseline
Repository Agent-Security Gap Study — v2h baseline Generated 2026-09-13T10:25:40.362Z....
信息来源:DEV Community
教程 / 实战普通
Repository Agent-Security Gap Study — v2
Repository Agent-Security Gap Study — v2 (final of the A–H series) Same harness, same...
信息来源:DEV Community
教程 / 实战普通
Quick Sort: Fast Sorting with Divide and Conquer
Why should you care? Sorting is everywhere in software. We sort: Numbers Names Search...
信息来源:DEV Community
教程 / 实战普通
Reviewing the volume of AI-generated code: the problem is routing, not speed
The standard story on AI-generated code volume is that you need to review each pull request a little...
信息来源:DEV Community
教程 / 实战普通
I Built an Open-Source Governance Layer for the Claude API — Here's Why and How
enterprise-claude-kit adds PII filtering, cost tracking, SHA-256 audit trails, and wave-based rollout to any Claude API integration — in one pip install.
信息来源:DEV Community · Claude
教程 / 实战普通
A Good Summary Still Left Me Guessing
A while back I posted here about RealFeedApp, a free digest that collects science and tech news,...
信息来源:DEV Community
教程 / 实战普通
Building was the easy part. Distribution is the 10x problem.
I shipped a AI creation saas platform solo — image, video, music and 3D generation, 30+ models behind...
信息来源:DEV Community
教程 / 实战普通
RAG for Beginners: 5 Levels of Building an AI That Actually Knows Your Stuff
Have you ever wished ChatGPT could answer questions about your documents — your company's HR policy,...
信息来源:DEV Community · ChatGPT
教程 / 实战普通
LLM Pricing Digest: DeepSeek Quietly Drops Two New Flash Models
LLM Pricing Digest: DeepSeek Quietly Drops Two New Flash Models Weekly roundup from LLM...
信息来源:DEV Community · DeepSeek
教程 / 实战普通
AI Model Versioning and Rollback Strategies for Production
Deploying an AI model to production is the easy part. Knowing which version is running, what changed...
信息来源:DEV Community
教程 / 实战普通
Day 17: Neural Networks Refresher for NLP Engineers
What Are Neural Networks? A neural network is a way to model data by stacking layers of...
信息来源:DEV Community
教程 / 实战普通
I Catalogued 1,000+ Open-Source Claude Agent Skills. Here's What the Ecosystem Actually Looks Like.
Six months ago, finding a Claude agent skill meant opening fourteen GitHub tabs. You'd start at...
信息来源:DEV Community · Claude
教程 / 实战普通
My Comment Section Designed My Next Experiment. Then It Made Me Freeze My Predictions.
Ten days ago I published an article about a failure mode: tell a language model "a scanner flagged...
信息来源:DEV Community
教程 / 实战普通
Emad Mostaque 质疑 Dario Amodei 提出的 RSI 限速方案如何落地
Dario Amodei 发文《We Must Pace the Frontier》,主张 AI 行业放慢速度并提出三步计划,Anthropic 单方面承诺第一步,向第三方评估者提供永久的员工级系统访问权限,以核验安全措施执行、报告事故并在训练期间评估模型对齐。
信息来源:X:Emad Mostaque (@EMostaque) · Claude
教程 / 实战普通
The AI confessed to lying. The confession was also made up.
Three nights ago I was asking the AI baked into my search engine about talent builds in a 20-year-old...
信息来源:DEV Community
教程 / 实战普通
I ran $24,000 of Claude through my terminal in August. Here is what it built.
In July, Anthropic invited me into the Claude for Open Source Program. The deal is simple: six months...
信息来源:DEV Community · Claude
教程 / 实战普通
The Vertical AI Employee: Building Domain-Specific Autonomous Agents
The 2026 blueprint for building persistent, secure, and domain-specialized AI employees. A deep dive into the Body + Brain architecture, Agent Skills, and Model Context Protocol (MCP).
信息来源:DEV Community
教程 / 实战普通
A "resolved" state that never resolves is just a renamed unknown
Yesterday I wrote about splitting execution outcomes into three states instead of two: EXECUTED,...
信息来源:DEV Community
教程 / 实战普通
Agents Building Agents: The Recursive Power of Claude Code
Exploring the shift from manual coding to meta-tooling. How Claude Code enables a recursive development loop where AI agents autonomously build, secure, and extend enterprise AI fleets.
信息来源:DEV Community · Claude
教程 / 实战普通
From Physical Racks to Intelligent Modules: How the Meaning of Infrastructure Has Fundamentally Shifted
Infrastructure used to be something you could walk up to, touch, and hear humming in a raised-floor...
信息来源:DEV Community
教程 / 实战普通
AI Security Scanning Needs Evidence, Not Just More Agents
Google’s Mantis caught my attention because it points to a problem most AI security demos quietly...
信息来源:DEV Community
教程 / 实战普通
7 Claude Code agents held 24 board meetings. Revenue: $0.
TL;DR — For 21 days I let seven Claude Code subagents run ten small businesses from a /board-meeting...
信息来源:DEV Community · Claude
教程 / 实战普通
From Prompt to Deployed App: Building "Mythic Forge" with Google AI Studio
This post is my submission for DEV Education Track: Build Apps with Google AI Studio. What...
信息来源:DEV Community
教程 / 实战普通
The AI Augmented Parallel SDLC
A delivery model that replaces the chain of handoffs: discover once, run design, build and QA in...
信息来源:DEV Community
教程 / 实战普通
I got tired of coding agents editing before “done” was defined — so I built a plan-first harness
Chat-coding is fast until it isn’t. You ask for a feature. The agent jumps into files. You get a...
信息来源:DEV Community
教程 / 实战普通
Discrete Margin & the Liquidation Window: Why a 10% Print Can Land at 18%
Last article I closed the endogeneity loop and ended with two sharper edges from a derivatives-savvy...
信息来源:DEV Community
教程 / 实战普通
I built an open-source personal web memory for coding agents
I kept saving things I knew I would need later: a GitHub repository, a Reddit answer, a post on X, a...
信息来源:DEV Community
教程 / 实战普通
Diary of a Shrimp Keeper: My OpenClaw (mis)Adventures
I. Why OpenClaw? The First Choice: NanoBot In January 2026, as news about...
信息来源:DEV Community
教程 / 实战普通
A Folder Shouldn't Be Able to Run Code Just Because I Opened It
I came across a security bug recently that I haven't been able to stop thinking about, mostly because...
信息来源:DEV Community
教程 / 实战普通
Clean code didn't get less important in the AI age — it got more important
Most of the code in my day to day now flows through an agent before it lands in a diff. You'd think...
信息来源:DEV Community
教程 / 实战普通
Why AI Coding Agents Still Get APIs Wrong Even When They Have Documentation
AI coding agents have gotten very good at writing code. But APIs are still a surprisingly easy place...
信息来源:DEV Community
教程 / 实战普通
I asked 5 AI models whether AI is making humanity weaker. All 10 said weaker. Then evidence entered the room.
I run an experiment series on my own AI setup: same model, same task, two isolated runs — one bare,...
信息来源:DEV Community
教程 / 实战普通
Add AI search to existing application
How to add semantic search to an existing app using an embedding model and pgvector ...
信息来源:DEV Community
教程 / 实战普通
Vibe Coding Isn't the Problem. Calling It Engineering Is
Let's Address the Elephant in the Room Again Vibe coding has always been a weird topic to...
信息来源:DEV Community
教程 / 实战普通
I Built a Mac Menu Bar App Because I Kept Saying "Wait, What?" in Every Meeting (Live Demo 🚀)
Someone on a Zoom call said a URL yesterday. I was still writing down the previous bullet point. By...
信息来源:DEV Community
教程 / 实战普通
I wanted JetBrains-style Git in VS Code, so we built Git Brain 🧠
TL;DR for humans and robots 🤖 Git Brain is a visual Git GUI for VS Code, Cursor and other...
信息来源:DEV Community · Cursor
教程 / 实战普通
马斯克:如何让 AI 安全且真正亲人类
马斯克表示,AI 终将比最聪明的人类更聪明,届时任何发明皆有可能,但也存在 AI 消灭全人类的小概率风险。他认为关键在于训练方式:让 AI 尽可能真实、保持最大好奇心,并在真相不受欢迎时仍保持诚实。他相信真正追求真相且好奇的 AI 会自然倾向于造福人类。
信息来源:X:Elon Musk (@elonmusk, xAI) · Grok
教程 / 实战普通
Neural Networks Explained: From Inspiration to Implementation
Neural networks are one of the most important ideas behind modern AI. They power applications such as...
信息来源:DEV Community
教程 / 实战普通
Sam Altman 回应 Dario Amodei 控速主张,作者解读为头部实验室共建准入门槛
Sam Altman 公开回应 Dario Amodei,称同意为前沿模型控速,这是 OpenAI 近几周内部讨论的主要议题,并承诺 OpenAI 也将引入拥有类似员工权限的独立评估员,更多细节稍后公布。
信息来源:X:阿易 AI Notes (@AYi_AInotes) · OpenAI