Research Scientist, AI Safety

Responsibilities

• Rigorously think through important open scientific questions that don’t have textbook answers.
• Design and build safety evaluations to understand AI capabilities at the frontier of biology.
• Collaborate closely with scientists across Biohub to build safeguards that are robust to potential misuse, enabling our AI systems to be used for maximum societal benefit.
• Gain direct access to a network of leading external experts and Biohub partners for collaboration and to inform your safety work.
• Monitor capability advancements and emerging technologies across AI and assess how these advances impact the robustness of our safeguards.
• Partner closely with colleagues to develop, implement, and evolve our frontier safety policy.
• Communicate our safety work to internal teams as well as external partners.

• PhD in computer science or a related field, or equivalent experience.
• Expertise in biological sciences.
• Significant experience building, using, and/or evaluating deep learning systems, ideally in research-driven environments.
• Strong engineering and scientific fundamentals.
• Comfort with ambiguity, rapid iteration, and loosely-defined research problems.
• Strong communication skills across technical and scientific audiences.
• Interest in shaping the frontier of safety for AI & biology; no prior safety experience required.