Spencer Kaplan

Overview

Spencer Kaplan is presented as a Yale PhD candidate who conducted two years of anthropological fieldwork in San Francisco’s AI research community.

Current Profile

Kaplan’s role in the source is interpretive rather than predictive. He reports how AI safety workers form and act on catastrophic-risk beliefs, connecting their choices to firsthand model work, Effective Altruism, and a Bay Area norm of personal agency while declining to validate any particular extinction estimate.

Key Characteristics

• Conducted two years of embedded fieldwork among San Francisco AI researchers.
• Observed online spaces, group houses, parties, reading groups, and salon-style discussions.
• Treats AI-risk belief as sincere enough to shape careers, savings, and ordinary life for some workers.
• Explains both public resignation and continued lab employment as possible impact-maximizing choices.
• Distinguishes the social practice of numerical risk estimation from the accuracy of any particular estimate.

Evidence

• Fieldwork basis: Inside the mindset of AI safety workers describes Kaplan’s two-year immersion in the community’s online and in-person life.
• Interpretive claim: Inside the mindset of AI safety workers attributes divergent stay-or-leave choices to a shared commitment to agency and expected impact.
• Probability boundary: Inside the mindset of AI safety workers says Kaplan learned probability assignment in an effective-altruism course but declined to judge a specific extinction forecast.

Qualifications

The supplied source is a short structured episode summary, not Kaplan’s dissertation, field notes, participant sample, or a technical evaluation of AI risk. It cannot establish prevalence across the AI industry, causal influence for every worker, or the accuracy of the beliefs described.

What Changed

• Created the profile from the Marketplace Tech interview summary.

Relationships

• Effective Altruism - movement and decision framework Kaplan identifies as influential among AI safety researchers.
• AI Safety Worker Culture - fieldwork community he interprets.
• Existential Risk Probability Estimation - decision practice he describes without validating a specific estimate.
• Jacob Coxon - public resignation example discussed in the episode.
• Marketplace Tech - program presenting his account.