The AI Edge · Watch · READ · Make
READ
A short, working reading list — documentation to build with, newsletters to stay current, and articles that shaped how I think about agentic AI. Read what shapes real thinking, not just the headlines.
Build with Claude
The primary sources. Start here when you're ready to actually build, not just read about building.
Claude Code overview
The official documentation for Claude Code — Anthropic's command-line agent for delegating real coding tasks. The canonical starting point for the agentic-coding workflow this whole curriculum builds toward.
Build with Claude
Anthropic's hub of guides and learning resources for building applications on Claude — prompting, tools, agents, and best practices, all from the source.
Claude Code repository
The public GitHub repo — install instructions, issues, and release notes. Worth following to see what's changing and to read real usage in the issue threads.
Stay current
Subscribe to one or two and let them filter the firehose for you.
Lenny's Newsletter — AI hub
Practical, product-minded writing on building and growing software, with a dedicated AI topic hub. Strong on how AI actually changes the way products get made — not just what's new.
The Briefing, by Martin Peers
Sharp daily analysis of the biggest stories in tech and AI from The Information's newsroom. Industry-insider perspective on the business and power dynamics behind the headlines.
Deeper dives
Longer reads that go past the surface — including a few I've curated myself.
drC's curated Medium list
My own Medium profile and reading list — the articles I've found worth keeping, updated as I go. A good place to check for what I'm currently reading and writing.
Model Context Protocol — intuitively and exhaustively explained
A thorough walkthrough of MCP, the open standard for connecting AI models to tools and data. If you've wondered how Claude reaches calendars, files, or APIs, this is the deep explanation.
Stop prompting, start designing: 5 agentic AI patterns that actually work
A practical move from one-off prompts to repeatable agentic patterns — the design-level thinking that separates a clever demo from something that holds up in production.
Agentic and multi-agentic AI
Traces "agentic" back to its roots — Bandura's psychology and the foundational multi-agent systems research — then connects that lineage to today's orchestration engines. Useful for understanding that agents aren't as new as the hype suggests.
The reading habit
Block thirty minutes a day for reading. Same time every day if you can. Read with your notebook open. Save the best articles to your Medium lists. Re-read what stuck a week later.
Most people read AI news. The people who actually understand the field read AI thinking.