Applications open for Cooperative AI PhD Fellowship

They are looking for early career researchers, who are enrolled or are about to be enrolled in a PhD programme at an accredited university (anywhere in the world), with a specific interest in multi-agent/cooperation problems involving AI systems. Your research proposal should aim to contribute to societally beneficial AI development.

The Cooperative AI Foundation is committed to the growth of a diverse and inclusive research community, and they especially welcome applications from underrepresented backgrounds. This is the first iteration of the fellowship, and they are still working out how they can best provide value to PhD students – if you apply now, you can help them and future generations of students.

Benefits

• They provide flexible funding adapted in duration and extent to the needs of each candidate, to give everyone an equal footing.
• If you already receive a stipend for your living or tuition expenses, they provide top-up funding up to this amount. The fellowship will be renewed each year for the duration of your PhD, conditional on you continuing to work on cooperative AI. In addition, they also provide:
  • Invitations and travel funding to CAIF’s events, including some that are private and/or aimed at senior researchers;
  • An additional platform for your research via CAIF’s website, events, social media, and other channels;
  • Opportunities to collaborate with CAIF’s researchers, advisors, and other organisations in their network, and to be involved in research initiatives;
  • A mini-grants program to meet your other needs (e.g. for travel to relevant events, minor research or equipment costs, family allowance – such as childcare costs for a child up to 4 years old, etc.);
  • Compute credits (matched to the needs of your research work) and (where possible) early access to the latest models and resources;
  • An annual well-being budget ($600 per year).

Funding Information

• They offer maximum funding of up to $40,000 per year to cover your living expenses (if tuition is required, they will cover tuition costs in addition).

Eligibility Criteria

• Applicants can be from any country or region.
• Applicants can apply to undertake their PhD in any subject at any department at an accredited university, as long as their research focuses on cooperative AI.
• Applicants with existing funding from other organisations are not prohibited from applying, though see also the question on “top-up” funding.
• Applicants can apply before or after they are accepted to a PhD program (including if they have already started their PhD), but fellowship awards are conditional on the applicant enrolling in a program in order to undertake the research outlined in their proposal.

Selection Criteria

• Their primary criterion is the potential to conduct impactful, high-quality cooperative AI research, both during your PhD and afterwards. As evidence of this potential, they evaluate applications on the strength of your research proposal as well as your experience, existing skill set, and career aspirations. Research proposals should:
  • Be focused on multi-agent/cooperation problems involving AI systems;
  • Be such that the results could be relevant for the most advanced AI systems, including future systems;
  • Aim to contribute in a major way to societally beneficial AI development.
• ‍Further instructions about the content of research proposals can be found on the application form. As with their regular grantmaking, the following criteria will be central in their evaluation of your research proposal:
  • Impact: If your research is successfully executed, how important would the results be from a societal perspective? Could this be among the most influential work in the field of cooperative AI right now?
  • Tractability/feasibility: Is your research likely to be successfully executed? Have you identified a tractable (set of) problem(s), and do you have the ability to carry it out? Please consider identifying any skills you might need to develop during your studies.
  • Neglectedness: If your research were not to go ahead, how likely is it that very similar work would happen soon anyway?‍
  • Risk: Is there a risk that your research might have harmful consequences, such as contributing to the development of more dangerous systems? Can you mitigate such risks?

Application Requirements

• You will be asked to submit a CV, a two-page research proposal, two reference letters, and to respond to a few short questions. Your data will be processed in accordance with their privacy policy.
• ‍Applications will be reviewed by CAIF staff as well as external researchers, where required, in two stages: screening and full review. In a small number of cases, they may invite you to a brief call. They hope to communicate final decisions by the end of November 2024.

For more information, visit Cooperative AI Foundation.