Every recommendation here comes from direct experience. I don't list things I haven't used. If something is missing, it's either because I haven't tested it enough, or I tested it and it didn't make the cut. New to AI agents? Start here first.

AI PLATFORMS — WHERE TO START
Claude PRIMARY

"My primary. Best reasoning, best tool use, best for structured work. Claude Code is how I build everything in this newsletter — every page, every agent, every tool. If you only try one, try this."

"Good for broad research and quick brainstorming. I still use it when I need a second opinion or want to stress-test an idea against a different model's reasoning. The free tier is genuinely useful."

Gemini

"Largest free context window. Useful for processing long documents — legal agreements, financial statements, SOPs — when you don't want to pay per token. The free tier handles 1M tokens."

"Replaced Google for 80% of my searches. Grounded answers with citations. When I need to research a competitor, check a regulation, or verify a claim, this is where I start. The Pro tier is worth it."

BUILDING BLOCKS — THE STACK
Claude Code ESSENTIAL

"CLI-native AI development. This is how every tool, every agent, every page on this site was built. You describe what you want, it writes the code, runs it, debugs it. Not optional for serious agent work."

"Anthropic's open standard for connecting AI to your tools. Think USB-C for AI integrations — one protocol, any data source. We run 8 MCP servers (Google Calendar, ElevenLabs, fal-ai, Canva, and more)."

Cursor

"Visual AI code editor. Good on-ramp if the command line feels intimidating. I started here before moving to Claude Code. Still useful for visual diffs and file navigation on large projects."

Zapier / Make

"No-code automation for connecting SaaS tools. Good for operators who aren't writing code yet. Connect Stripe to Slack, or Square to Google Sheets. Start here, graduate to agents when you outgrow it."

READING — GO DEEPER

"Primary source. When Claude gets better at something, this is where they explain why. Dense reading, but worth it. Their papers on constitutional AI and tool use shaped how I design agent guardrails."

"Best independent voice on LLM tooling. Practical, skeptical, detailed. He builds things, breaks things, and writes about both. If you only follow one person in this space, follow Simon."

"Technical AI conversations for practitioners. The guests are builders, not pundits. Episodes with AI engineers at Anthropic, OpenAI, and startups are consistently the highest-signal content in my feed."

"Product and ops perspective on AI adoption. Good for the business side of the equation — how to think about AI in your product roadmap, team structure, and go-to-market. Complements the technical sources."

COMMUNITIES — FIND YOUR PEOPLE

"Technical, practitioner-focused. Where builders hang out. The signal-to-noise ratio is high because the bar for entry is actual engineering work, not opinions. Good place to ask hard questions."

"Filter for AI threads. Brutal but honest feedback. I post FFN essays here because the comments are more valuable than any focus group. Thick skin required. Worth it."

Local Meetups

"Show up to whatever AI meetup exists in your city. The best partnerships, the best deals, and the best insights happen face-to-face. In Honolulu, that's the Hawaii AI Summit. Find yours."

FROM THE WORKSHOP — BUILT ON FFN PRINCIPLES
Agent Team Composer

"Interactive tool from Field Note #001. Design your own multi-agent team — pick the roles, define the boundaries, see the orchestration pattern. Built to make the essay actionable."

Coming with Essay #001 launch

48-Hour Decision Audit

"Classify any decision as one-way or two-way door, get a deadline, stop overthinking. Based on the Framework essay on decision velocity. The most-used tool in my own operating system."

Coming with Framework #001

Get field notes from the frontier.

Two essays a month. AI-first strategy, real numbers, tools you can ship today.