赫尔辛基大学 + MinnaLearnUniversity of Helsinki + MinnaLearn

无需数学或编程基础，用生活化方式理解 AI 能做什么、不能做什么，以及它如何影响社会。

A no-math, no-code introduction to what AI can and cannot do, and how it shapes everyday life.

李宏毅 · 台湾大学 / YouTubeHung-yi Lee · National Taiwan University / YouTube

20 堂生动讲座串起提示技巧、LLM 训练、智能体、Transformer、可解释性与 AI 安全。

Twenty lively lectures spanning prompting, LLM training, agents, Transformers, interpretability, and AI safety.

IBM SkillsBuild

覆盖 AI 核心概念、伦理、提示写作与跨行业都能使用的基础职场技能。

Covers core AI concepts, ethics, prompt writing, and foundational workplace skills for any field.

Google for Developers

通过交互练习掌握回归、分类、过拟合、神经网络与嵌入，并延伸到 LLM、生产 ML 与公平性。

Interactive practice in regression, classification, overfitting, neural nets, embeddings, LLMs, production ML, and fairness.

Kaggle · Google

在浏览器里完成 7 节动手课：数据探索、模型验证、欠拟合/过拟合与随机森林。

Seven browser-based lessons covering data exploration, validation, under/overfitting, and random forests.

李宏毅 · 台湾大学 / YouTubeHung-yi Lee · National Taiwan University / YouTube

31 堂大学课程从深度学习一路讲到 CNN、Transformer、GAN、BERT/GPT、强化学习与可解释 ML。

Thirty-one university lectures spanning deep learning, CNNs, Transformers, GANs, BERT/GPT, RL, and explainable ML.

fast.ai · Jeremy Howard

以代码和项目为先，用 PyTorch 与 fastai 构建视觉、NLP、表格与推荐模型，第二课即做出可部署模型。

A code-first path through vision, NLP, tabular, and recommender models, with a deployable model by lesson two.

DeepLearning.AI + OpenAI

学习提示工程最佳实践，并用 LLM API 完成摘要、推断、改写、扩写与自定义聊天机器人。

Prompting best practices plus hands-on summarization, inference, transformation, expansion, and chatbot building.

Hugging Face

从 NLP 基础进入前沿大模型，学习 Transformer、微调、量化与部署及 Hugging Face 工具链。

Bridges NLP foundations to modern LLMs through Transformers, fine-tuning, quantization, deployment, and the Hugging Face stack.

Anthropic Academy

用 4D 框架——委派、描述、辨别、勤勉——学习如何高效、合乎伦理且安全地与 AI 协作。

Anthropic's 4D framework—Delegation, Description, Discernment, and Diligence—for effective, ethical AI collaboration.

DeepLearning.AI · Andrew Ng

在 AI 助手陪伴下从零学习 Python，通过食谱生成器、智能待办与数据分析器建立编程直觉。

Learn Python from zero with an AI assistant while building recipe generators, smart to-do lists, and data analyzers.

DeepLearning.AI + Replit

与 AI 编程智能体协作，用 PRD、线框图与有效调试完成并部署两个真实 Web 应用。

Collaborate with an AI coding agent, using PRDs, wireframes, and debugging to build and deploy two real web apps.

DeepLearning.AI + Anthropic

学习代码库探索、项目记忆、MCP 与并行子智能体，并完成 RAG 聊天、数据看板和设计转应用项目。

Agentic coding through codebase exploration, project memory, MCP, parallel agents, and three applied projects.

Microsoft

7 章动手实践，围绕一个示例项目学习运行编程智能体、审核计划、查看差异与验证测试。

Seven hands-on chapters on running coding agents, reviewing plans and diffs, and validating tests in one sample project.

OpenAI Academy

建立提示与复核习惯，设计可重复的 AI 工作流，并学习如何用智能体推进实际任务。

Build habits for prompting and review, create repeatable AI workflows, and direct work with agents.

Google Cloud · Google Skills

用短课理解生成式 AI 与传统 ML 的差异、常见用途，以及 Google 的生成式 AI 工具。

A compact introduction to generative AI, how it differs from traditional ML, common uses, and Google's tools.

Anthropic Academy

聚焦如何把人类判断与 AI 智能体的执行力组合起来，在团队中更有效地分工、协作与复核。

How to combine human judgment with agent execution for stronger delegation, collaboration, and review in teams.