Idea: work closely with extremely talented folks on specific in-house projects, with an eye towards spinning out a new startup/org
As opposed to an “incubator” (batch of 5-20 people overlapping, byo idea), or “fellowship” (MATS-like research + mentor). Maybe could be shortened to “resident”.
Why I’m excited for “Founder in Residence”
- We have some track record of having done well with this approach (The Curve)
- And maybe soon: Mox, Falcon Fund, Manifest.
- I currently already have too many projects to manage
- I have a bunch of specific ideas on things to build (some of which are high-context or work well within Mox & Manifund), and I want someone founder-shaped to take them
- Also more generally, the AI safety & EA ecosystem has lots of ideas but is bad at execution
- There hasn’t been a single successful incubator for AI safety
- … so maybe the approach of incubator is bad for AI safety?
Project ideas
<see EA/AIS Software Incubator>. Maybe also “AI rights fieldbuilding”.
Questions
- What is the right fiscal umbrella structure for this kind of thing?
- Venture? Traditional SAAS/startup VC is optimized for fast growth
- Alphabet/Google model (?) — cash cow (search) pays for a bunch of related bets
- University model (?)
- CG model — many funds
- See also the failure of “Effective Ventures”
- See also https://nadia.xyz/idea-machines
- What are the most successful orgs started by FIRs?
- What kinds of people make for good FIRs?
- What things do FIRs want from the role?
- cofounders?
- ideas?
- money?
- advice, mentorship?
- “hero license”?
- How much explicit programming to provide, vs “just wander around and do stuff”
- Forprofit vs nonprofit
- Difference in fundraising (angels then vcs, vs donors then funds)
- Difference in mentality (”provide value, scale” vs “pitch to funders”)
- Why didn’t Izzy/Aethra succeed? What’s up with Constellation’s incubator?
Appendix
Notes from Vishal & Ishani May 13, 2026: