How to apply for Research Collaborators

About University of California, Center for Human-Compatible Artificial Intelligence

The Center for Human-Compatible Artificial Intelligence (CHAI) at UC Berkeley is a leading academic research group founded by Stuart Russell, dedicated to ensuring AI systems are provably beneficial to humanity. Unlike industry labs focused on product deployment, CHAI emphasizes foundational, interdisciplinary research that combines computer science, mathematics, and the humanities to align AI with human values. Working here means contributing to a global movement to steer AI development toward safety and ethical robustness.

About the role

As a Research Collaborator, you will join a distributed, interdisciplinary team to advance CHAI's mission by contributing to research projects that span AI safety, value alignment, and human-compatible design. This contract role is ideal for scholars and researchers who want to collaborate remotely on cutting-edge problems, from formalizing human preferences to analyzing the societal impacts of AI. Your work will directly inform the development of provably beneficial AI systems and help shape the field's ethical foundations.

A typical day

A typical day might involve asynchronous collaboration with a global team: reviewing recent literature, coding or writing up analyses for a value-alignment project, and participating in a virtual lab meeting to discuss progress. You might also attend a weekly seminar (if Berkeley-affiliated) or engage in deep work on a philosophical or mathematical problem. The schedule is flexible but requires proactive communication and self-motivation.

Who University of California, Center for Human-Compatible Artificial Intelligence is looking for

• An interdisciplinary researcher with a strong background in at least one of CHAI's core areas (computer science, mathematics, robotics, cognitive psychology) and a demonstrated interest in human value systems (moral philosophy, sociology, law).
• A collaborative self-starter who thrives in a remote, asynchronous environment and can engage with a global network of scholars.
• For undergraduates: a clear aspiration to pursue graduate study at UC Berkeley, with a desire to build a research portfolio aligned with CHAI's mission.
• For UC Berkeley affiliates: availability to participate in weekly research seminars and contribute to on-campus intellectual life.

Tips for this application

• Highlight any prior work or coursework that bridges technical AI research with human values—e.g., a project on preference learning, fairness in ML, or normative ethics applied to technology.
• Explicitly state your interest in CHAI's specific research agenda (e.g., inverse reinforcement learning, cooperative AI, value learning) and mention relevant papers by CHAI faculty like Stuart Russell.
• If you're an undergraduate, articulate your intention to apply to UC Berkeley graduate programs and how this role would prepare you; if you're a Berkeley affiliate, mention your ability to attend weekly seminars.
• Demonstrate remote collaboration skills by providing examples of successful distributed teamwork, such as open-source contributions or virtual research projects.
• Tailor your CV to show interdisciplinary breadth: list publications, talks, or coursework in both technical and humanistic fields, and use your cover letter to weave a coherent narrative.

What to cover in your cover letter

["Your specific motivation for contributing to CHAI's mission of provably beneficial AI, referencing particular research projects or papers from the center.", 'Concrete examples of interdisciplinary collaboration, especially where you integrated technical methods with insights from philosophy, sociology, or law.', "How your background aligns with the role's requirements, including your interest in human value systems and, if applicable, your plans for graduate study at UC Berkeley or participation in seminars.", 'Your ability to work effectively in a remote, global team, with evidence of self-direction and asynchronous communication skills.']

Research before applying

• Read CHAI's foundational papers, especially those by Stuart Russell (e.g., 'Human Compatible' book, 'Research Priorities for Robust and Beneficial AI'), to understand their technical and philosophical stance.
• Explore current CHAI projects listed on their website, such as inverse reward design, assistance games, or value alignment, and identify which ones align with your expertise.
• Look into the backgrounds of CHAI's core faculty and affiliates to understand the interdisciplinary culture and potential collaborators.
• Familiarize yourself with UC Berkeley's graduate programs in relevant departments (EECS, Philosophy, etc.) if you're an undergraduate aiming to study there.

Likely interview topics

Based on the job description, expect questions about:

• Discuss a research project where you addressed a problem at the intersection of AI and human values—what was your approach, and what did you learn?
• How would you formalize or operationalize a concept like 'human values' in a computational framework? What are the challenges?
• Which CHAI research paper or project resonates most with you, and how would you extend it?
• Describe a time you collaborated across disciplines. How did you bridge different vocabularies and methodologies?
• What are the biggest risks of advanced AI, and how does CHAI's approach of 'provably beneficial AI' address them compared to other safety paradigms?

Common mistakes to avoid

• Submitting a generic application that doesn't mention CHAI's specific mission or research areas—this suggests a lack of genuine interest in human-compatible AI.
• Overemphasizing technical skills without connecting them to human values or interdisciplinary collaboration, which is central to this role.
• Neglecting to address the remote nature of the role; failing to demonstrate ability to work independently and communicate effectively online can be a dealbreaker.