Anthropic Founders Used Stuffed Animals for Management
Daniela Amodei and Holden Karnofsky adopted toy personas to navigate managerial challenges while at OpenAI.
Updated on Sept. 30, 2026 in Artificial Intelligence

Live Poll
Do you trust executives who use unconventional or whimsical methods to make business decisions?
Between 2018 and 2020, Daniela Amodei and Holden Karnofsky employed a personal advisory council of stuffed animals during their time at OpenAI. The pair used these specific toy personas to model different personality types and resolve complex internal leadership problems.
Why it matters
The practice highlights the unconventional internal dynamics that existed among the group of founders before they departed OpenAI to form Anthropic. These behavioral modeling strategies allowed the team to process interpersonal conflicts that eventually led them to launch their own competing firm.
The startup eventually grew beyond its initial 50-person employee vision to include 8 co-founders. Analysts have speculated that the company could achieve a future valuation as high as $2 trillion in an IPO.
The players
Daniela Amodei
She is a co-founder of the AI safety company Anthropic and previously worked at OpenAI.
Holden Karnofsky
He is a prominent figure in the effective altruism movement and helped establish Anthropic.
OpenAI
This is an artificial intelligence research organization responsible for developing large language models like GPT.
Anthropic
This is an AI safety and research company focused on building steerable and reliable artificial intelligence systems.
The details
The founders maintained separate Slack channels and held private meetings to organize their departure from OpenAI, using the stuffed animals to simulate management advice. Personas included Jinji, a cat that suggested canceling meetings, Maura, a detail-oriented bear, and Beary Bonds, an empathetic panda.
Timeline
Daniela Amodei and Holden Karnofsky were married in 2017.
Daniela Amodei began her tenure at OpenAI in 2018.
Dario Amodei initiated recruitment efforts in 2020.
The team left OpenAI to establish Anthropic in December 2020.
Amodei publicly disclosed the advisory council practice in 2023.
The Tech Race
The unconventional management strategies of the founders occurred against the backdrop of the intense development of large language models. These behavioral habits mark a distinct cultural departure from the rigid corporate structures typically associated with the rapid scaling of artificial intelligence.
Users interacting with AI models today benefit from the organizational independence that resulted from these early management practices. The shift in leadership culture directly influenced the development priorities of current mainstream AI tools available to the public.
The takeaway
Using symbolic proxies can provide a necessary psychological buffer for leaders dealing with intense professional conflict. Implementing role-playing exercises may help individuals gain perspective on complex interpersonal problems in high-pressure work environments.
Further reading
For more on the industry's evolution, explore Artificial Intelligence.
Source note: This article includes information reported by New York Post.
Live Poll
Do you trust executives who use unconventional or whimsical methods to make business decisions?










