Technical Governance Researcher

About the Company

The Machine Intelligence Research Institute (MIRI) is a nonprofit based in Berkeley, California, focused on reducing existential risks from the transition to smarter-than-human AI. While historically focused on technical alignment, MIRI has expanded its focus towards communication and AI governance.

Responsibilities

• Conduct threat modeling to understand how AI systems could cause large-scale harm
• Respond to government Requests for Comment (RFCs)
• Research risk management practices in other industries and apply them to AI
• Design and implement evaluations of AI models to demonstrate failure modes
• Prepare and present informative briefings to policymakers
• Analyze government or developer AI policy documents for limitations
• Design new AI policies and standards addressing current limitations
• Maintain focus on solutions that scale to smarter-than-human intelligence

Requirements

• Strong background in AI Safety
• Experience or familiarity in compute governance, AI policy, or as an AI safety generalist
• Ability to communicate effectively with external stakeholders and policymakers
• Strong written communication skills for research and reports
• High degree of agency, conscientiousness, and generative thinking
• Ability to learn quickly and work on underspecified tasks

Preferred Qualifications

• Research or engineering experience focused on frontier AI models or the AI tech stack
• Experience creating model evaluations or benchmarking AI hardware
• Experience conducting scaling law experiments or empirical work

Benefits

• Health, dental, and vision insurance (Kaiser and Blue Shield)
• "No vacation policy" allowing staff to take time when needed in coordination with managers
• Potential visa sponsorship for promising candidates