The network around Anthropic, METR, and Effective Altruism is a tight cluster of people who overlapped at OpenAI, lived in the same Bay Area EA houses, married each other, and moved through the same Oxford and philanthropy pipelines. Here is how the main people and organizations connect.
Core organizations (the “flow”)
GiveWell / Open Philanthropy / Coefficient Giving**: Founded and led by Holden Karnofsky. This is the main EA grantmaking machine that funded AI-safety work for years. Ajeya Cotra spent nearly a decade there. Dustin Moskovitz’s Good Ventures money flows through it.
Oxford EA cluster**: Future of Humanity Institute (FHI), Centre for Effective Altruism, Global Priorities Institute, Centre on Long-Term Risk (CLR). This is the academic feeder. Hjalmar Wijk, Megan Kinniment, and Pip Arnott all came through Oxford + FHI/CLR.
OpenAI (2016–2021)**: Shared workplace for Dario Amodei (VP Research), Daniela Amodei (VP Safety & Policy), Paul Christiano (alignment lead, co-inventor of RLHF), and Beth Barnes.
Alignment Research Center (ARC)**: Founded 2021 by Paul Christiano after he left OpenAI. Incubated the evaluations team.
ARC Evals → METR**: Started 2022 inside ARC under Beth Barnes. Spun out as the independent 501(c)(3) METR in December 2023. METR now does third-party capability and threat evaluations for Anthropic, OpenAI, Google, Meta, etc.
Person-by-person origins and ties
This will be a thread
1.🧵👇
Anthropic CEO Dario Amodei's handpicked AI watchdog has deep ties to woke Effective Altruism movement: 'a complete joke'
trib.al/IuA3uHW