Research Scientist — Alignment
Lead research on reward modeling, preference learning, and constitutional AI. PhD required.
Apply →Research Engineer — Interpretability
Build tools and infrastructure for mechanistic interpretability. Strong ML systems background required.
Apply →Research Scientist — Safety Evaluations
Design and run red-teaming evaluations. Background in adversarial ML or security research.
Apply →Research Scientist — Scaling
Study scaling laws, compute-optimal training, and emergent capabilities. Strong math background.
Apply →Senior Software Engineer — ML Platform
Build infrastructure for model training and evaluation pipelines. Distributed systems experience.
Apply →Policy Researcher — AI Governance
Research and communicate on AI policy, safety standards, and governance. Technology policy background.
Apply →How we work.
Not slogans. Real operating principles that shape every decision.
- 01
We argue in writing. Evidence before intuition.
- 02
We change our minds when the data requires it.
- 03
Every person has outsized impact. No bureaucracy.
- 04
Mission alignment is required, not optional.
- 05
Long time horizons change what we build.
- 06
Comfort with uncertainty is a core skill here.