AI-For-Beginners

A 12-week, 24-lesson AI curriculum from Microsoft, with lessons, quizzes, and labs. It starts with symbolic AI, moves through neural networks and deep learning, and covers ethics alongside frameworks like TensorFlow and PyTorch.

The repository is a full curriculum: 12 weeks, 24 lessons, with practical lessons, quizzes, and labs. It is pitched as beginner friendly and mentions ethics in AI up front. The framework story runs through TensorFlow and PyTorch, and the whole thing is distributed as notebooks plus reading material.

The syllabus starts with the symbolic approach to AI, knowledge representation and reasoning, then moves into neural networks and deep learning illustrated with TensorFlow and PyTorch code. Computer vision and natural language processing get their own sections, and less common approaches like genetic algorithms and multi-agent systems are included too. Business uses of AI, classic machine learning, cognitive services, and the mathematics behind deep learning all appear in the learning list, with pointers to Microsoft Learn paths and a recommended textbook by Goodfellow, Bengio, and Courville.

Each lesson bundles pre-reading material and an executable Jupyter notebook, often in both PyTorch and TensorFlow versions, where the notebook carries substantial theory as well as code. Labs exist for some topics so learners can apply what they covered. Quizzes live in a quiz-app folder that can run locally or deploy to Azure, and several sections link out to Microsoft Learn modules.

For complete beginners, a short list of examples is meant to be run before the full curriculum. Hello AI World is a first pattern-recognition program, Simple Neural Network builds one from scratch, Image Classifier walks through image classification with detailed comments, and Text Sentiment analyzes positive and negative text. They are designed to build intuition before the structured course starts.

More than 50 language translations come bundled, which inflates the download size, so a sparse checkout command is documented for cloning without them. Learners are invited to a Discord server, and product feedback routes to an Azure AI Foundry forum. The README credits Dmitry Soshnikov as primary author, Jen Looper as editor, Tomomi Imura for sketchnotes, and Lateefah Bello for the quizzes, along with core contributors.

At roughly 65,000 stars, the curriculum has found its audience. It is translated into over 50 languages, ships beginner examples for quick starts, and names its authors and editor at the bottom of the README.

• Source 1 · Project repository
• Source 2 · README