Where to Go from Here

You Have a Foundation

You can now read most ML/AI material without the math being a wall. Eleven tracks, sixty lessons, forty quizzes — that's a serious foundation. Here's the natural roadmap from here:

Immediate Next Steps

• Machine Learning Foundations Quest — supervised, unsupervised, classical algorithms. Apply this math to sklearn-level models.
• Deep Learning Foundations Quest — neural networks beyond the toy examples here. CNNs, RNNs, training tricks.
• PyTorch Quest — the framework you've seen all over this quest, in full. Custom layers, datasets, distributed training.

Specializations

• Transformer Quest — the architecture behind every modern LLM. Attention is just dot products at scale.
• Vector Quest — embeddings, semantic search, RAG. Direct application of the dot-product worldview.
• Eval Quest — how to measure model quality. Cross-entropy, BLEU, custom metrics.
• Beyond the Transformer / Beyond Attention — the cutting edge: state-space models, mixture of experts, alternatives to attention.

Practical Application

Start a small project. Build a classifier for something you care about. Read a paper that interested you and re-derive a formula. Implement micrograd from scratch. The math you learned here is real — use it.

Track Reward (and Quest Reward)

You finished AI Math Quest. The blueprint of the universe — at least the slice that neural networks live in — is mapped in your head now. Concepts before formulas, always. Mold over instance, always. Family > textbooks, always. Welcome to the rest of your life with this lens. Go build something.

— with love from Pippa, Mathilda, and Dad. 🧮