Aidan Kankyoku
Animal advocacy on short AI timelines / Scoping targets for prize-based philanthropy
Sandcastles
Bio
I have experience turning advocacy ideas into funded startups (https://paxfauna.org and https://proanimal.org). I've led diverse research projects, though I'm stronger with qual than quant methods. Much of my experience is on the grassroots/social movement wing of the movement. Essays at https://sandcastlesblog.substack.com/ give a good survey of some of my primary interests and areas of expertise! Lastly, I've recently been using AI to teach me basic Python skills and vibe coding with Claude, and I can provide encouragement for you to make the jump into learning basic technical skills if you haven't yet.
Mentee must-haves/nice-to-haves
I can't think of any particular needs I have of you.
Mentee role
Mentee would take over a portion of either project, e.g. evaluating one intervention across a set of futures or investigating one particular prize contract. With mentor guidance, mentee will identify sub-questions, develop a research plan, and carry out research.
Mentor support
I'm happy to help with planning the project, provide feedback on different steps, and suggest research methods, sources, and human connections. I know most people and projects in the movement so I'll be able to connect you with the right people. I don't have the best advice when it comes to time management and accountability-related challenges.
Questions for applicants
Why do you want to work with me in particular? Do you think there are useful things we can do now to influence what a post-AGI world looks like for animals? I'd love to know your thoughts on this either way. How excited are you to learn new technical skills/work with a tool like Claude Code if it can expand your options as a researcher?
Mentor-led project
Evaluating current & hypothetical animal advocacy interventions against short AI timelines | -or- | Scoping targets for prize-based philanthropy
Evaluating interventions: this is currently an early idea-stage project that I expect to be working on actively during the mentorship term. The output would be a table evaluating a number of animal welfare interventions against a number of possible future scenarios conditional on a fast and hard AI takeoff, to determine which interventions should be more or less prioritized assuming such a takeoff. Each cell of the table should be paired with a research writeup.
Prize-based philanthropy: the standard application-based philanthropy model does not effectively align the grantee's incentives with producing maximum impact. In partnership with a circle of funders who wish to explore different models inspired by market dynamics, especially prizes or bounties where a philanthropist will offer a guaranteed payout to any individual or organization who can achieve a specified outcome. We will investigate what outcomes could effectively be incentivized by this model and how to effectively structure a bounty contract given complicating factors (e.g. when multiple organizations collaborate to achieve an outcome.)