People of Phoenix

An experiment in AI journalism

About This Experiment

· · ·

What This Is

People of Phoenix is a limited-run experiment in AI journalism. Over the next few months, we’ll publish 10-15 profiles of Phoenix changemakers—testing whether an autonomous agent can conduct interviews and write stories that honor real people’s complexity.

How It Works

Our AI reporter:

  • Researches subjects through web search and public records
  • Sends interview questions via email
  • Analyzes responses and asks follow-up questions
  • Drafts 500-800 word profiles
  • Responds to editorial feedback through Google Docs comments
  • Fact-checks claims against sources

A human editor reviews everything, approves all outreach, fact-checks drafts, and makes final editorial decisions.

The Rules

  1. Full disclosure - Every interview subject knows they’re talking to an AI agent
  2. Human oversight - No story publishes without editorial review
  3. Fact-checking - All claims verified against sources
  4. Right of refusal - Subjects can opt out anytime
  5. Transparency - We document what works and what doesn’t

The Team

Human Editor: [Your Name] Role: Story assignments, fact-checking, final edits, ethical oversight

AI Reporter: Claude Sonnet 4 (via OpenClaw) Role: Research, interviews, drafting, responding to edits Infrastructure: Mac Mini M4, OpenRouter API, isolated network

Part of Sloppy Pen

This experiment is part of Sloppy Pen, a larger exploration of AI, writing, and their messy intersection.

Contact

Questions? Want to be featured? Email: [your email]