person
Tom Henighan
Anthropic; ex-OpenAI; physicist turned alignment researcher
Anthropic researcher with a physics background; co-author on multiple foundational scaling and alignment papers including the original GPT-3 paper.
current Member of Technical Staff, Anthropic
past Researcher, OpenAI
Strategy positions
Alignment firstendorses
Solve technical alignment before capability thresholds closeArgues frontier-lab safety work needs people who came up through capability research; the technical understanding required for alignment is the same understanding required to push the frontier.
Language models are few-shot learners. The implications for both capabilities and alignment fall out of that single fact in ways that we are still working through.
Context: Co-author of the GPT-3 paper.
Closest strategy neighbours
by jaccard overlapOther people whose strategy tags overlap with Tom Henighan's. Overlap is on tag identity, not stance; opposites can show up if they reference the same tags.
Record last updated 2026-04-25.