A curated map for developers lost in AI's endless forked road

What it does

This repository is a living collection of resources for learning artificial intelligence—books, courses, papers, videos, tools, models, and notes. The author, a former games industry developer, treats it as a journal rather than a curriculum: here’s what I used, what helped, and where I got stuck. The README opens with a brutally concise TL;DR for the time-starved: one book, one video, one course, then expands into categorized deep-dives.

The interesting bit

The curation carries credibility precisely because it admits uncertainty. The author notes, “Do I know AI? Not really.” This isn’t a professor’s syllabus—it’s a practitioner who learned by building, applying the same “pick a game and make it” philosophy that works in game development to the far messier domain of AI tooling. The result is opinionated without being preachy.

Key highlights

• Tiered depth: opens with a 3-item “limited time” survival kit, then scales to 5+ books, 5+ courses, 8+ videos, tool matrices, and model hubs
• Video curation is unusually sharp: prioritizes primary sources—Andrej Karpathy’s deep dives, 3Blue1Brown’s visual explanations, MIT bootcamp recordings—over tutorial farm content
• Tool taxonomy by function: separates core frameworks (Hugging Face, Ollama), developer tools (Gradio, Colab), visualization (LLM Visualization, custom embeddings viewer), and compute platforms (Lambda)
• Acknowledges ecosystem complexity: explicitly states that “understanding all the tools… can sometimes feel harder than understanding AI itself”
• Extends beyond Python: includes a C++ neural-network-from-scratch weekend project for the masochistically curious

Caveats

• The articles/blogs section appears truncated in the source; recency and completeness there are unclear
• Python library coverage is a standard laundry list (Pandas, PyTorch, etc.) without the same contextual commentary given to other sections
• No explicit filtering for free vs. paid resources; some recommendations (O’Reilly books, Lambda compute) carry real costs

Verdict

Worth bookmarking if you’re a developer who learns by doing and wants a curated on-ramp rather than another overwhelming “awesome-list.” Skip it if you need structured progression or academic rigor—the author explicitly disclaims both.