What it does This is a hand-picked index of machine learning resources—articles, courses, videos, books, and demo projects—tagged with emoji difficulty levels. It was built to support The Coding Train’s YouTube videos and an ITP course, but functions as a standalone roadmap for self-learners.

The interesting bit The emoji taxonomy is genuinely useful: :rainbow: for creative approaches, :bowtie: for beginners, :sweat_smile: for intermediate, :godmode: for advanced. Most curation dumps are chaos; this one at least admits when something will hurt your brain.

Key highlights

• Heavy tilt toward creative and visual ML (Neural Aesthetic, char-rnn, pix2pix-style image generation)
• Includes browser-based demos you can break immediately (K-nearest, Q-learning games, self-driving car sim)
• Surprisingly good video section: full MIT AI lectures, DeepMind’s David Silver RL course, Sherbrooke neural nets class
• Actually mentions Big O notation—rare honesty about the fundamentals ML tutorials skip
• Covers practical entry points (fast.ai, Google’s crash course) alongside the weird stuff (evolving soft-bodied creatures)

Caveats

• No code in the repo itself; it’s purely a link collection
• Some links are circa 2015–2017 and may rot; the README doesn’t note last-checked dates
• “Examples and experiments” in the description is aspirational—most content is third-party

Verdict Worth bookmarking if you’re a creative coder or educator building an ML curriculum. Skip it if you want runnable code or a structured course; this is a syllabus, not a textbook.