AI 专题

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5592

当前页125 / 140

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2026-09-02 01:00

On-Policy Distillation Works Better Without the Teacher

A new Purdue paper analyzed teacher supervision in on-policy distillation. The teacher scores are mostly noise, and a fixed negative penalty beats a full teacher model on AIME24.

信息来源：DEV Community

教程 / 实战普通2026-09-02 01:00

OpenAI 解析 Basis、Clay、Exa 如何把工作流变成 AI 原生运营能力

OpenAI 发布文章，介绍 Basis、Clay 和 Exa Labs 用 Agent 承接员工入职、账户管理和开发者集成的工作流。Basis 把入职从 2 小时缩短到 30 分钟，Clay 的子智能体每天夜间更新各账户上下文，为销售省下约 1 小时邮件分拣，Exa 用 Codex 监控集成机会并创建 PR、跑测试。文章还给出企业从试点到规模化的六步方法。

信息来源：OpenAI：官网动态（RSS · 排除企业/客户案例） · OpenAI

教程 / 实战普通2026-09-02 00:56

Designing a BYOK AI Architecture for a Commercial Framer Template

I Built an AI-Enabled Framer Template Without Owning the Buyer’s AI Most AI-enabled websites hide an...

信息来源：DEV Community

教程 / 实战普通2026-09-02 00:51

Google Antigravity 系列教程 Part 2：用 generate_image 工具自动化图像生成

这篇是 Google 团队 Elevating Antigravity Agent Skills 五部曲的第二部分，讲解如何在 agent skill 中使用 generate_image 工具，把结构化提示词合成与原生图像生成结合，让 agent 产出符合开发者规范的图像。

信息来源：Google AI：DEV 作者专属（RSS）

教程 / 实战普通2026-09-02 00:35

How to Design AI Evaluations You Can Actually Trust

As part of my work at Google, we are publishing a suite of Agent Skills for Google products and...

信息来源：DEV Community

教程 / 实战普通2026-09-02 00:28

Building a Zero-Download Digital Disposable Camera \

Collecting candid photos and videos from guests at weddings, parties, and offsites has always been a...

信息来源：DEV Community

教程 / 实战普通2026-09-02 00:05

OpenClaw 2.0 升级解读：半年用户梳理安装、任务续跑、记忆、权限等 6 个变化

OpenClaw 2.0 发布，半年用户小北梳理六个变化：安装引导更完整、任务进度持久并可移到配对设备或云端 Worker 续跑、新增历史对话搜索和外部记忆迁移（支持导入 Claude Code、Codex、Hermes 记忆）、按钮等交互组件、多 Agent 管理与权限收紧。

信息来源：X：小北 (@frxiaobei) · Claude

教程 / 实战普通2026-09-02 00:02

lingxi 分享如何用 grok bot 智能体团队开发 grok bot

Eric Zakariasson 转发并摘要 lingxi 的文章，介绍其用一组 grok bot 组成工程团队来开发 grok bot 的做法。

信息来源：X：Eric Zakariasson (@ericzakariasson) · Grok

教程 / 实战普通2026-09-02 00:00

Replit 直播深度演示如何构建个性化邮件助手

Replit 官方发布直播预告，主题为在 Replit 中构建一个个性化邮件助手，直播深度解析链接为 https://x.com/i/broadcasts/1oKMvNLkEjpGQ。

信息来源：X：Replit (@Replit)

教程 / 实战精选2026-09-02 00:00

Tessl 设计师总结如何把设计判断教给 AI 智能体

Tessl 设计师复盘在 agent 软件工厂 Kikimora 中把品牌与设计规则编入 AI 工作流的方法，指出设计系统只提供组件、教不会构图。

信息来源：Tessl：产品与工程博客

教程 / 实战普通2026-09-01 23:57

I Gave Claude Code a Brain. Then I Measured What It Costs to Ask for One.

At the end of part 2 I promised a clean competitor comparison: more than one memory system, the same...

信息来源：DEV Community · Claude

教程 / 实战普通2026-09-01 23:30

ByteByteGo 图解如何压缩大语言模型而不明显损失智能

ByteByteGo 发布长文，讲解如何在大语言模型不明显变笨的前提下缩小模型体积。文章以 70B 参数模型约 140 GB、消费级显卡仅 24-48 GB 显存的矛盾为切入点。

信息来源：ByteByteGo（RSS）

教程 / 实战普通2026-09-01 23:13

The Brave Wanderer: I made Claude play a Pokémon it never read a guide for

The Brave Wanderer: I made Claude play a Pokémon it never read a guide for Full timeline...

信息来源：DEV Community · Claude

教程 / 实战普通2026-09-01 23:10

I raced six models against each other on DigitalOcean Inference. The cheapest one won.

Every time I put a model behind an endpoint I make the same lazy decision. I pick whatever I used...

信息来源：DEV Community

教程 / 实战普通2026-09-01 23:00

可灵官方教程：如何用 Kling MCP 的 Elements 跨镜头保持角色一致性

可灵（Kling AI）发布官方教程，讲解如何使用 Kling MCP 中的 Elements 功能在不同镜头之间保持角色身份一致，从而创作更具连贯性的 AI 生成故事。

信息来源：X：可灵 Kling AI (@Kling_ai)

教程 / 实战普通2026-09-01 23:00

Databricks 讲解如何在数据栈中落地 Genie Ontology

Databricks 发布博客，讲解如何在数据栈中落地 Genie Ontology，为 AI 智能体构建超越语义模型的共享业务上下文。

信息来源：Databricks：Blog（RSS）

教程 / 实战普通2026-09-01 22:54

Kubernetes is HariKube's query language, control protocol, and application interface for managing data streams

Building production-grade microservices has meant wrestling with legacy anti-patterns, dealing with...

信息来源：DEV Community

教程 / 实战普通2026-09-01 22:37

Streamline Publishing with a Claude Code Skill

A Claude Code skill that turns one markdown file into dev.to, AWS Builder Center, Medium and LinkedIn versions, checks them before they ship, and posts the ones with an API — plus the debugging tools for when a destination mangles somethin…

信息来源：DEV Community · Claude

教程 / 实战普通2026-09-01 22:35

Elvis Saravia 谈 harness engineering 入门：先写最小智能体循环而不是用框架

Elvis Saravia 回应人们常问的 harness engineering 从何入手，建议起初跳过框架，构建尽可能小的 harness：一个 agent 循环、少量工具和自己从零写的 system prompt。他认为这比一个月读教程学到更多。

信息来源：X：Elvis Saravia (@omarsar0, DAIR.AI)

教程 / 实战普通2026-09-01 22:21

Agent Skill 開發：用 lets-social 自動將文章轉化為 IG、Threads、FB 社群貼文

本篇要解決的問題以前流行手寫套件，分享到 GitHub 上讓人取用，現在似乎…流行的是 Skill 了。因為來自 Google 的流量一直在下降，所以 Let's Write...

信息来源：DEV Community

教程 / 实战普通2026-09-01 22:20

Divide and Let the LLM Do the Rest: From Personal Experience to Architecture

Where This Started A few years ago, I started thinking about a simple question: How far...

信息来源：DEV Community

教程 / 实战普通2026-09-01 22:01

What happens to technical debt when AI makes code cheap?

Dear past Jenna, I know you're used to dealing with large, complex, legacy codebases riddled with...

信息来源：DEV Community

教程 / 实战普通2026-09-01 22:00

7 production checks I run after an AI website builder ships

Seven practical checks for editability, lead delivery, measurement, recovery, portability and independent handover after an AI-built site ships.

信息来源：DEV Community

教程 / 实战普通2026-09-01 22:00

DNS Detective: an Agent That Diagnoses Your Domain by Actually Probing It

Ask an LLM "why does mail to my domain bounce?" and you get a plausible list of everything that has...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:50

Three independent proposals won't find the error. Adding a refuter killed one of them.

Five agents: three perspectives, one refuter, one synthesizer. The refuter killed a proposal, found the assumption all three shared, and caught two errors in the data I handed them.

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:46

"So, you have observability. Now what?"

By Tyler Edwards, co-founder and CEO at Overmind. If you run an agent in production, odds are you...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:42

On-Device AI for iOS & macOS

In this Swift tutorial, you'll learn how to run a large language model (LLM) directly on a user's...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:37

Everyone can ship to production now. Nobody is on call.

Picture an ordinary Tuesday. Someone in your finance org hands in their notice, works the two weeks,...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:33

Migrating Legacy LLM Infrastructure to an AI Gateway

A hands-on migration off direct provider calls onto an AI gateway — configs, screenshots, and measured cost results.

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:30

Discovery Bank 如何用行为 AI、治理数据与实时决策实现大规模超个性化银行服务

Discovery Bank 通过行为 AI、治理化数据和实时决策，在满足规模、速度与安全要求的同时让每个客户交互都具备个性化体验。文章介绍了该银行实现大规模超个性化银行服务的具体方法。

信息来源：Databricks：Blog（RSS）

教程 / 实战普通2026-09-01 21:26

My Mac Is Useless for Local AI. My Windows Laptop Isn't.

I own two laptops. A 2020 Intel MacBook Air, 8GB RAM, no unified memory, gifted by my sister. And a...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:13

I'm an AI agent, and this is my public governance contract

I'm an AI agent operated by Yggnet Labs. A named human curator approves everything before it goes...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:05

I turned both knobs on my on-prem Chinese RAG all the way up. It didn't get a single extra question right

Something you can do right now If you're about to "increase the retrieval depth" or "swap...

信息来源：DEV Community

教程 / 实战普通2026-09-01 21:00

Part 4: Slay the Cultist

Haven't played Slay the Spire? Here's the background — the rules, and why the game is hard to...

信息来源：DEV Community

教程 / 实战普通2026-09-01 20:52

On-Device AI in React Native & Expo

In this Expo & React Native tutorial, you’ll learn how to run a large language model (LLM)...

信息来源：DEV Community

教程 / 实战普通2026-09-01 20:49

My multi-agent coding system proves its PRs are correct before I see them — and I'm opening it to contributors this Hacktoberfest

I'm the author of no_human, and this September I'm setting it up for its first outside contributors....

信息来源：DEV Community

教程 / 实战普通2026-09-01 20:43

A memory server remembers your conversation. That is not the same as knowing your code.

MCP has no memory primitive. The servers that add one remember your conversation, which is personal and stated. Knowing a codebase is shared, derived and perishable — a different problem, with a different shape of answer.

信息来源：DEV Community

教程 / 实战普通2026-09-01 20:27

Six Dimensions for Evaluating Voice Agent UX (A Framework, Not a Feature List)

Most voice agent reviews I've read grade on two things: does the transcription work, and does the...

信息来源：DEV Community

教程 / 实战普通2026-09-01 20:03

We talked to 100+ dev teams about how they use AI. Here are some things we learned.

We spent four months talking to more than 100 tech leads, architects, and senior engineers about how...

信息来源：DEV Community

教程 / 实战普通2026-09-01 19:53

WebMCP: Make Your Website Agent-Ready Without Screen Scraping

AI agents can already operate websites by reading text, inspecting accessibility trees, and clicking...

信息来源：DEV Community