The author, a former employee at Google DeepMind, details their reasons for leaving, primarily stemming from the company's support of the immigration enforcement supply chain and the signing of an AI deal with the Pentagon that lacked restrictions against killer robots or mass surveillance. The author attempted to influence Google's decisions by appealing to senior figures like Jeff Dean and Demis Hassabis, and by proposing a detailed framework for ethical AI contracts. Despite these efforts and assurances from senior management, Google proceeded with the deal.

Key figures such as Stuart Russell, a prominent AI researcher, and Jeff Dean, Google's Chief Scientist, were approached for support, but their actions or inactions are described as falling short of the author's expectations. Russell's organization failed to make a promised statement, and Dean, despite signing a pledge against killer robots and co-signing an amicus brief supporting Anthropic, did not appear to leverage his influence to stop Google's own military AI deal. The author's comprehensive proposal for contract language was ultimately ignored.

Feeling unable to reconcile their ethical concerns with the company's actions, particularly after Google's contract restrictions were deemed weaker than OpenAI's, the author resigned. The essay reflects on the failure of powerful individuals and institutions to uphold AI ethics promises under pressure, highlighting the personal cost of such a decision and the broader implications for AI safety and governance.