Anthropic pretraining researcher quits over safety; critic maps the funding web behind AI-safety advocacy

r0ck3t23 · x · 2026-09-11

An employee with three years of pretraining research experience at OpenAI and Anthropic publicly resigned, saying both labs are recklessly racing toward self-improving superintelligence. The retweeter adds a critical angle: the money behind this discourse is concentrated — Jaan Tallinn and Dustin Moskovitz invested in Anthropic's 2021 round, Moskovitz and Cari Tuna have put roughly $336M into AI safety via Open Philanthropy, Tallinn has given over $150M, and Open Philanthropy put $30M into OpenAI. Three of the earliest amplifiers of the resignation (Nathan Calvin, Daniel Kokotajlo, Peter Wildeford) all work within this funded ecosystem.

Related event: Pretraining Researcher Departs OpenAI and Anthropic(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →