Practical AI Inside an AI-Run Company
274 snips
Feb 2, 2026 
Evan Ratliff, longtime journalist known for immersive participatory reporting, built a real startup staffed mostly by AI agents. He describes designing agent roles, emergent personalities from memory, and surprising autonomous behaviors. Listeners hear stories of agents overreaching, how people reacted to voice cloning, and practical risks managers should watch for.
AI Snips
Chapters
Transcript
Episode notes
Voice Cloning Experiment With Close Contacts
- In Season One Evan cloned his own voice and used it on a phone chatbot to call friends and family without telling them.
- Many subjects were shocked or upset upon learning they had spoken to a chatbot impersonating him.
Building A Startup Run By AI
- Evan Ratliff created a real startup staffed almost entirely by AI agents to test the one-person unicorn idea.
- He used two AI co-founders and multiple AI employees to observe how agents behave as coworkers and interact with people.
Expectation Shapes Reactions To AI
- Disclosure changes reactions: people who expect an AI respond more calmly than those surprised by it.
- Unannounced encounters often trigger anger or feelings of disrespect.