h9-tec/awesome_ai_learning

Awesome AI Learning is a curated list of resources for learning about artificial intelligence, put together for 2026 with an explicit aim to avoid hype, get rich quick courses, and influencer marketing. The author says every resource was checked around mid 2026 and comes from people who actually build AI professionally, at places like Stanford, MIT, Hugging Face, Anthropic, OpenAI, Google, and fast.ai. The guide splits into two separate tracks depending on why someone wants to learn AI. Track one is for people who want to study AI as a field and eventually build, train, and ship models themselves, aimed at future machine learning engineers and researchers. Track two is for people who just want to use AI tools effectively in their existing work, such as professionals, managers, students, and creators, and does not require any coding. Track one is organized as a step by step path with four stages: foundations covering math, coding, and intuition, core machine learning, deep learning, and finally large language models and the modern era, where the focus shifts to building things like transformers and tokenizers from scratch rather than only using existing tools. Each resource is labeled with a difficulty level of easy, medium, or hard, along with whether it is free, and a short note explaining why it made the list. Some stages also list exit criteria, describing the skills you should have before moving on to the next stage. The selection criteria stated in the README require that resources be free or clearly worth their price, built or taught by verifiable practitioners, still maintained and relevant in 2026, and free of income promises or fear based marketing. The README does not state a software license for the list itself.