7 Resources to Learn AI Seriously

Post by Alexandre Zajac — curated AI learning path

The 7 Resources

0. Latent Space

Newsletter + podcast by swyx & Alessio. They interview the people who actually built the thing, not people who write about it. Their Discord Paper Club gets you reading papers you’d otherwise skip. → latent.space

1. Simon Willison’s Blog

Every time a model drops, he’s testing it same-day with runnable code. No hype. Just “here’s what it actually does.” → simonwillison.net

2. “AI Engineering” by Chip Huyen

Not another prompting thread — the systems-level reference. Eval design, RAG tradeoffs, when fine-tuning is actually worth it. → Book link

3. Karpathy’s “Neural Networks: Zero to Hero”

Built in 2022, still the best antidote to prompt-tip content. Backprop and GPT from scratch. → YouTube series

4. Eugene Yan’s Blog

Production war stories on evals and recsys. Postmortems, not hot takes. → eugeneyan.com/writing/

5. Hamel Husain’s Blog

The one resource that teaches you how to actually know if your LLM app works. Evals on steroids. → hamel.dev/#blog

6. Hugging Face Courses + Blog

Free, written by the people who maintain the libraries you’re already importing. → huggingface.co/blog

Key Takeaway

“Interview the people who actually built the thing, not people who write about it.” “Postmortems, not hot takes.” “No hype. Just here’s what it actually does.”