A reading list that knows the field is moving too fast

What it does This is a curated list of books, courses, papers, frameworks, and tools for building and shipping AI systems. It is organized around practical engineering concerns — RAG, agents, evals, deployment — rather than pure research or startup press releases.

The interesting bit The curation has a point of view. The author explicitly favors “deep, durable knowledge” over trend-chasing, and includes a personal note advising you to start with simple LLM calls rather than immediately adopting heavy frameworks. The “Agents” section is framed as “harnesses that turn LLMs into autonomous workers” — a useful reminder that the model is swappable and the orchestration layer is what you’re actually buying or building.

Key highlights

• Heavyweight reading list mixing canonical texts (Russell & Norvig, Sutton & Barto, Goodfellow et al.) with recent practical books by Chip Huyen, Sebastian Raschka, and others
• Framework coverage ranges from minimalist learning tools (PocketFlow, ~100 lines) to production orchestration (LangGraph, CrewAI, AutoGen)
• Coding agent section is unusually comprehensive — 17 entries including terminal-native tools, IDE extensions, and local-first options
• Model comparison section covers language, image, video, and audio with brief, opinionated “best for” labels
• Includes live benchmark and pricing aggregators (OpenRouter, LMArena, Artificial Analysis)

Caveats

• No stated criteria for inclusion or update cadence; some “best for” claims are subjective one-liners without supporting evidence
• The README is a single flat file with no search, tagging, or filtering — discoverability degrades as the list grows
• Several sections are link-dense with minimal annotation; you’ll need to click through to assess quality yourself

Verdict Worth bookmarking if you’re an AI engineer trying to stay current without drowning in noise. Less useful if you want rigorous, peer-reviewed curation or interactive exploration — this is a well-read practitioner’s personal index, not a systematic review.