AGI Institutions Research Fellowship

We need more research on institutions¹ that work in a world with powerful AI. Courts, contracts, markets and elections were all built for a world of human participants who are relatively slow, and sensitive to things like norms and social sanctions. Autonomous agents break these assumptions, and may cause significant strain on these institutions as they start participating in them.

We better have good proposals ready for the moment when politicians and policymakers start looking for alternatives.

For more on this, read the Pax Machina announcement post.

You won’t arrive with your own project. Researchers in our network will pitch you projects, and help shape the one you choose together with you.

Each project tackles one of the most pressing problems in AGI institutions. For example, you might research new market mechanisms that delegate the right kind of economic activity to agents, new corporate forms that keep organizations accountable to richer mandates, or new adjudication mechanisms that can handle rising caseloads from AI.

Good work will likely include simulation or experiments with participants (e.g. on Prolific). We’ll cover compute and study costs.

We expect a research paper or article at the end of the fellowship.

• Mentorship and access to world-class experts. We’ve curated a research network of ~100 or so of the most promising researchers working on these questions, and will pair you up with a relevant researcher.
• We’ll help drive adoption of the most promising research output, either through publishing ideas in Pax Machina, by finding the right co-authors for your paper, or by connecting ideas with builders.
• We’ll pay you $15k/mo for the duration of the fellowship.

We’re looking for talented early-career researchers from relevant fields (economics, law, political science, cognitive science, computer science, philosophy) who are excited to work on AGI institutions with the best people in the world.

A good candidate is:

• High energy and self-directed. You can quickly and effectively scope and execute on research projects.
• Motivated by effecting real change in the world. We’re at a point in history where the right research in this field could have an outsized impact on the future.
• Somewhat familiar with the relevant work in the field. See here for what kind of work we consider to be relevant for AGI institutions.

Fill out the form below, describing some problem or grid cell from agi-institutions.org that excites you, and a high-level idea for a research project around it (note that we will help you shape your actual project together with relevant mentors!)

We’ll select candidates on a rolling basis.

Questions? Email research@meaningalignment.org.