Google DeepMind is looking to solve the problem of ensuring the safety of its AI systems, specifically in the context of Chemical, Biological, Radiological and Nuclear (CBRN) risks, by developing and implementing evaluations and mitigations.
Requirements
- Proven ability to write clean, maintainable, and efficient Python.
- Knowledge of concepts in mathematics, statistics, and machine learning needed for understanding cutting-edge research in AI, CBRN-related modeling, and risk assessment.
- Familiarity with relevant datasets, benchmarks, or evaluation methodologies for CBRN risks in AI.
- Experience developing or working with safety-critical systems.
- Experience with data analysis tools and libraries for large-scale evaluation.
- Knowledge of international and national regulations or guidelines related to CBRN materials or technologies.
- Ability to present and explain technical results clearly to non-experts and leadership stakeholders.
Responsibilities
- Design and develop robust evaluations to test potential CBRN risks arising from cutting-edge AI models with strong capabilities in the physical sciences.
- Develop and maintain infrastructure for these evaluations.
- Run these evaluations prior to releases for new AI models.
- Resolve potential CBRN risks with state-of-the-art mitigations.
- Clearly communicate results to decision-makers.
- Collaborate with subject matter experts in CBRN, AI policy and ethics, model development, and scientific research.
- Develop and maintain an understanding of trends in AI development, CBRN threat landscapes, governance, and relevant sociotechnical research to inform the design of new evaluations.
Other
- Bachelor's degree in a technical subject (e.g., computer science, engineering, machine learning, mathematics, physics, statistics) or a relevant scientific field with strong computational experience.
- Master's or PhD in a field relevant to CBRN risk assessment or AI safety, such as chemical, biological, or nuclear engineering.
- Ability to work on projects with many stakeholders, including scientific experts, policy teams, and engineers.
- Commitment to equal employment opportunities regardless of sex, race, religion or belief, ethnic or national origin, disability, age, citizenship, marital, domestic or civil partnership status, sexual orientation, gender identity, pregnancy, or related condition.
- Background check performed by a third party acting on Google DeepMind's behalf.
