# Survival and Flourishing Fund (SFF) **Draft profile.** **[[Jaan Tallinn]]**'s primary philanthropic vehicle for existential-risk and AI-safety grantmaking. Founded **2019** with an initial ~$2M from the **Berkeley Existential Risk Initiative** (BERI), itself Tallinn-funded. SFF has since become **one of the two most consequential x-risk funders in the world** (the other being [[Open Philanthropy]]), distributing approximately **$141M** between 2019 and 2025 — the bulk drawn from Tallinn's **~$900M Skype-founder fortune**. The 2025 round alone totaled **$34.33M**, with **86% directed to AI safety**. ## What Makes SFF Unusual SFF is structurally different from standard foundations in two important ways: ### 1. It is a "virtual fund" SFF doesn't hold endowment in the traditional sense. Instead, it operates as a **decision-and-distribution mechanism** that organizes the **application, evaluation, and grant-routing process**. The actual capital flows from Tallinn (and a small number of other donors) into the recommended grantees, with SFF acting as the coordinating layer. ### 2. The S-Process SFF's signature innovation is the **"S-process"** — a formal mathematical procedure for allocating grants: - **6–12 Recommenders** are selected for each round. - Each Recommender expresses their preferences as a **utility function** over the applicant pool. - An **algorithm aggregates the utility functions** into a final allocation. - Critically, the S-process **favors projects with enthusiastic champions over consensus picks** — a deliberate design choice meant to avoid the "compromise candidate" trap of grant committees. The S-process produces grant lists that **no single Recommender would have produced individually**, and it is one of the few formal grantmaking algorithms with published methodology. Median S-process grant: ~$100K. ## Grant Programs SFF operates three parallel grant streams: ### Speculation Grants - ~35 **Speculation Grantors**, each with a budget allocation. - Decisions can be made as quickly as **one week** after application. - Funded primarily by direct Tallinn donations. - Designed for **fast, high-conviction bets** on early-stage projects or individuals. ### S-Process Grants - The main twice-yearly grant rounds using the S-process algorithm. - 6–12 Recommenders, ~$100K median grant. - Funds organizations rather than individuals. ### Initiative Committee Grants - A small group (Tallinn + 2–5 anonymous SFF Advisors) **makes grants without applications**. - Used for opportunities where no formal application would be appropriate. ## Focus Areas - **AI safety / alignment** — the dominant category. 86% of 2025 distributions. - **AI governance** — policy and institutional work. - **Biosecurity** — pandemic preparedness, dual-use biology governance. - **Institutional resilience** — broader civilizational-survival infrastructure. - **Long-term-flourishing research** — academic and applied work on long-run trajectory. ## Why It Matters in This KB ### 1. It is Jaan Tallinn's primary instrument of x-risk movement-building Where [[Open Philanthropy]] is Moskovitz/Tuna's vehicle, SFF is **Tallinn's**. The two operations are not coordinated but their grant portfolios substantially overlap — they fund the same [[MIRI]] / [[CSER]] / [[FLI]] / academic-AI-safety ecosystem. **Together they constitute the majority of global x-risk philanthropy.** ### 2. The S-Process is a methodological innovation worth tracking SFF's algorithmic-grantmaking methodology is one of the few **public, formalized** alternatives to traditional foundation review processes. It has been studied by other funders (including elements of [[Open Philanthropy]]'s GiveWell-origin approach) and is a notable artifact of the broader **rationalist/[[Effective Altruism]]** community's interest in formalizing judgment. ### 3. It is a critical capital-flow node in the [[TESCREAL]] institutional stack SFF money flows to [[MIRI]] (founded by [[Eliezer Yudkowsky]]), [[Future of Life Institute]] (Tallinn co-founded), [[Centre for the Study of Existential Risk]] (Tallinn co-founded), [[Global Priorities Institute]], and dozens of smaller AI-safety operations. Tracing **who SFF funds** is one of the cleanest ways to identify the **active operational core of the x-risk movement** in any given year. ### 4. It demonstrates the donor-name-attached funding model Unlike Open Philanthropy (which is institutionally branded), SFF is **explicitly Tallinn's vehicle** — published donor lists, named advisors, public methodology. This is a notable contrast with the opacity of [[Marble Freedom Trust]] or [[DonorsTrust]] on the right. ## Critiques - **Network capture** — SFF's grantees disproportionately come from the same rationalist/EA community Tallinn personally identifies with. Critics argue the S-process formalism does not eliminate the network-selection effect at the Recommender stage. - **AI-safety concentration** — 86% of 2025 funding to AI safety is a major concentration bet that critics argue crowds out other x-risk categories (climate, biosecurity, governance). - **TESCREAL ideology critique** — SFF is one of the operational engines of the [[TESCREAL]] worldview that [[Émile Torres]] and [[Timnit Gebru]] critique structurally: concentrating decision-making in a small techno-elite about what futures are worth pursuing. - **Opacity at the Initiative Committee level** — the "2–5 anonymous voters" alongside Tallinn is the one place SFF retreats from its general transparency posture. ## Draft Notes / To Expand - Add detailed grant tables for major SFF rounds (2019–2026) - Document the S-process mathematical specification - Map the overlap with [[Open Philanthropy]] grants in any given year - Trace Tallinn's other philanthropic vehicles (the Estonian foundation, MetaCulture, etc.) - Detail [[Berkeley Existential Risk Initiative]] (BERI) as the prehistory of SFF - Investigate which Speculation Grantors / Recommenders have served across multiple rounds - Document the [[FTX]] / [[Sam Bankman-Fried]] collapse's effect on SFF's prominence (SBF was historically the major rival EA funder) ## See Also - [[Jaan Tallinn]] — Founder and primary donor - [[MIRI]] — Major recurring grantee - [[Future of Life Institute]] — Tallinn co-founded, recurring grantee - [[Centre for the Study of Existential Risk]] — Tallinn co-founded, recurring grantee - [[Open Philanthropy]] — Adjacent x-risk megafunder - [[Berkeley Existential Risk Initiative]] (currently unlinked) — Prehistory - [[TESCREAL]] — Adjacent ideological bundle - [[Longtermism]] — Adjacent ideology - [[Existential Risk]] — Adjacent concept - [[Philanthropy Ecosystem]] — Money-flow MOC