A 6-week online working group for people in AI governance, starting 11 January 2027
SAGE is a small online working group at the Center for Reducing Suffering (CRS) on how AI governance could reduce s-risk (risk of suffering on an astronomical scale). It’s for people with AI governance experience and people from s-risk research, and the purpose is to identify and prioritize open questions, and to identify potential AI governance interventions to reduce s-risk.
Key details
- Dates: 6 weeks, Mon 11 Jan to Fri 19 Feb 2027, plus up to two optional sessions in the two weeks after.
- Time: about 3 hours a week, with one 90-minute video call in a small group plus short readings or writing.
- Call times: we’ll set them after selection, to fit most of the people we accept. We can’t promise the timing will work out for everyone we’d like to accept.
- Facilitators: David Veldran and Alistair Stewart (CRS).
- Cost: free.
- Deadline: Sun 22 Nov 2026.
Who it’s for
- People who work on AI governance or policy, for example in government, a think tank, a lab’s policy team, or research on these. We’ll also take people from s-risk research who want to work on governance questions.
- You don’t need a philosophy background or to have worked on s-risk, but please read this intro first. The readings in the first two weeks cover what you need.
- It’s not an introduction to AI governance. The discussions assume you know the main institutions and debates.
- We’ll be selective and interview the strongest applicants before offering spots, so the discussions and what comes out of them stay strong.
What we’ll do (rough structure, subject to change)
- Weeks 1 and 2: short readings on s-risk and AI governance. Everyone proposes open questions, then we rank them together and discuss where we disagree.
- Weeks 3 to 5: one main question a week, with a short written reflection before each call.
- Week 6: we revisit the priorities, and each person names a next step or project idea (if they want).
- Calls start with reading and commenting on each other’s writing.
- We suggest a weekly question and reading; groups can change them.
- Optional extras: a joint session with the other cohort, and one with a guest on space governance.
Possible outcomes
- A prioritized list of open questions and potential interventions.
- Input on what CRS works on next, since we’ll use both cohorts’ rankings to inform our AI governance research priorities.
- Feedback on a next step or project idea, if you want it.
What happens with the results
After week 6, we combine both cohorts’ rankings and main disagreements, ask other s-risk and AI governance researchers for feedback, and use the output to inform what CRS works on next. Participants get the combined results.
How to apply
- Fill in the application form by Sun 22 Nov. It takes about 30 to 40 minutes and includes questions about your role, two short written answers (100 and about 200 words), and your availability between 11 Jan and 19 Feb.
- We’ll invite shortlisted applicants to a short video interview between 23 Nov and 6 Dec, and send decisions by Fri 11 Dec.
- Before the start, everyone agrees to the commitment and conduct expectations set out in a short doc.
Questions? Email programs@reducingsuffering.org.