Location: San Francisco, CA (on-site 5 days/week)
About The Center for AI Safety (CAIS)
The Center for AI Safety (CAIS) is a leading research and advocacy organization focused on mitigating societal-scale risks from AI. Some of our past achievements include: releasing the most widely used measure of AI capabilities used by all major AI companies, running a large compute cluster to facilitate AI safety research which has been cited over 16,000 times, and publishing a global statement on AI Risk signed by Geoffrey Hinton, Yoshua Bengio and top AI CEOs.
What distinguishes us is what we choose to work on. Our work is aimed at reducing real-world risks from advanced AI systems. We deliberately pursue research directions that the field is not yet paying attention to, and we move on once the rest of the field catches up. Our focus is on problems that are both highly important and highly neglected and our track record is built on getting to them first.
- In 2022–2023, we focused on AI honesty, robustness, transparency, and trojan/backdoor behaviors.
- In 2023–2024, we turned to malicious use and weaponization capabilities, introducing the first state-of-the-art benchmarks for measuring it.
- More recently, we've been working on AI value systems and the functional well-being of AI systems.
This is a research philosophy as opposed to a fixed agenda: we go where the important, unworked problems are. Because our work tends to be timely and to open up territory rather than crowd into it, our papers have repeatedly gone on to become widely cited and to set the standard for underexplored research areas. Our work is regularly used by AI safety institutes and frontier AI labs, and they have shaped real policy outcomes, including being presented directly to senators and policymakers.
The Role
You will act as a senior researcher who can unblock hard technical problems, a people manager who can develop and hold a team accountable, and a translator who bridges the Director's strategic direction and the team's day-to-day execution. The best person for this role combines research taste with the operational discipline to make a team reliably deliver.
Our work centers on empirical deep learning research with large language models and/or multimodal models.
What You'll Do
- Solve hard problems and unblock the team. Step in when researchers get stuck. Diagnose what's going wrong in an experiment or a research direction, propose a path forward, and get people moving again. You've been senior long enough to have seen these failure modes before.
- Manage research performance. Continuously evaluate the quality and pace of the team's work. Give honest, well-calibrated feedback, coach researchers toward higher standards, and surface who is delivering and who needs support.
- Set and maintain priorities. Translate the Research Director's vision into concrete priorities for the team, and keep the team's work aligned to them. Push back on low-value side quests and scattered directions; make sure effort flows to what matters most.
- Drive planning and accountability. Own project management for the team. Make sure tasks have clear owners and deadlines, that work is actually completed at the quality bar, and that slippage and dropped work are caught and corrected rather than quietly tolerated.
- Build a strong team culture. Shape team norms, resolve conflict, support morale and engagement, run an effective meeting cadence, and keep information flowing well across the team.
- Write and contribute to papers, and help hire researchers and interns as the team grows.
What We're Looking For
- Strong empirical deep learning research experience. You have personally led empirical research on LLMs or deep learning systems, with published work at top ML venues to show for it.
- Research taste. You can tell good research directions from bad ones, detect when an argument or result doesn't hold up, and know what to prioritize. Concretely, that means you:
- Can get up to speed quickly on a literature and track the latest developments.
- Can propose new, promising experiments and subdirections (often drawing on philosophy, concepts from other disciplines, and AI safety context).
- Can reason precisely about subtle conceptual distinctions—the kind of verbal and philosophical clarity needed to handle concepts like AI deception or AI wellbeing rigorously.
- Have a track record of strong published research.
- Research execution. You can validate experiments, reproduce or pressure-test results, and find the flaws in more junior researchers' work.
- People management. Ability to manage performance, coach people to improve, and build healthy team culture and collaboration norms.
- Project management. Organized, and able to keep work on track and people accountable. Prior project-management experience helps but isn't essential if you're naturally orderly and conscientious.
- The basics. Sharp, conscientious, and a strong communicator who can translate effectively between leadership and the team. A genuine collaborator—low ego and motivated by the mission of making AI safe.
Not sure you're a "manager"? Apply anyway. If you're a PhD student or someone with a strong research record who hasn't formally managed a team but is conscientious, organized, and has mentored or guided people before, this is a profile we're genuinely excited about. We weight research ability most heavily, and we'd rather find a great researcher we can grow into this role than someone with a management title and weaker research instincts.
Nice to Have
- Significant AI safety context and familiarity with the field's open problems.
- Prior experience collaborating with our Director or others on the team.
Salary Range
$170,000 - $260,000 annually, depending on experience.
Benefits
- Health insurance for you and your dependents
- 401K plan + 4% matching
- Unlimited PTO
- Lunch and dinner at the office
- Annual Professional Development Stipend
- Access to some of the top talent working on technical and conceptual research in AI safety