And what should Mox focus on this year?

Strategic visions

What would a thriving AI safety scene in SF look like? Some possibilities:

1. Connect lab people and AI safety people
   • professional events: happy hours, workshops, conferences
   • social/fun events (eg Joel’s karaoke)
2. Build an independent counterweight to labs
   • protest organizing, independent think tank/research hub
   • support independent orgs with a great office space & community
3. Fieldbuild AI safety, increasing available talent
   • fellowships, training programs
4. Incubate new orgs and fields
   • eg AI rights, sentience, welfare
   • eg xrisk grassroots advocacy, but good
   • eg better futures, AI for social coordination/human reasoning
5. others?

Specific things we could do

1. Longterm programs (fellowships, incubators)
   1. support existing programs
   2. run our own programs (?)
2. Events (happy hours, talks, retreats, conferences)
   1. including events run outside of Mox??
3. Offices
   1. Unlike Constellation, Mox doesn’t have a high density of AIS-y offices. We could try recruiting some of these. (partly, these orgs are currently in Berkeley)
   2. How much value add do we provide over WeWork, or DIY office?
4. Individual researchers
   1. Or people in between roles, looking for a place to hang out
   2. or “nights & weekends” space for eg lab employees. (this was one of the original hopes for Mox, but we’re currently not doing a great job of it imo)
5. others?

Other questions

• Should we move? Or expand to a second location?
  • Our current location at 1680 Mission St. is large, provisioned, relatively cheap, and fairly central/easy to get to; but it’s also in a rundown area, and not walking distance to many labs
  • Currently, METR is thinking about moving into offices to be very close to major labs (in FiDi). We could collab with them, or open our own events space around there
• Who would be a good person to lead SF AI safety?
  • See also: Role: SF AI safety organizer — as written it’s scoped to a “community organizer” role, but maybe this is the wrong approach
• How much should we try to “climb the prestige ladder” eg with premier orgs and researchers, vs build out a robust warm open community? (How much are these at odds?)

See also: https://manifund.org/projects/mox-2026-fundraiser for a 2025 summary of our work