# Future of Life Institute The **Future of Life Institute (FLI)** is a **non-profit research and advocacy organization** founded in 2014 in the Boston-Cambridge area by MIT physicist **Max Tegmark**, [[Jaan Tallinn]], philosopher Anthony Aguirre, and co-founders Viktoriya Krakovna and Meia Chita-Tegmark. FLI's stated mission is to reduce global catastrophic and existential risks from advanced technology, with particular focus on **artificial intelligence**, **biotechnology**, **nuclear weapons**, and **climate change**. Across its decade of operation the organization has hosted the 2017 Asilomar Conference that produced the **Asilomar AI Principles**, convened an advisory board that has included Stephen Hawking, Morgan Freeman, and a long list of prominent scientists, and — most visibly — **authored the March 2023 "Pause Giant AI Experiments" open letter** that was signed by Elon Musk, Steve Wozniak, [[Yoshua Bengio]], and over 1,000 other scientists and technology figures. FLI matters for the KB for three reasons. First, it is one of the **most politically visible organizations in the AI safety / existential risk ecosystem** — the March 2023 pause letter put AI-doom arguments on front pages worldwide and converted what had been an intramovement debate into mainstream political content. Second, it is the **principal point of overlap between the academic AI safety research community and the tech-industry figures who have backed the movement with significant capital** — Tegmark's MIT academic credibility, Tallinn's funding, Elon Musk's donations, and board members like Stephen Hawking made FLI the most mainstream-accessible face of a movement otherwise often dismissed as Bay Area eccentricity. Third, its **grantmaking operations** — $10+ million in grants to AI safety research since 2015 — have helped fund the technical AI alignment field and a range of adjacent organizations, giving FLI a concrete distributional role alongside its advocacy work. ## Founding (2014) FLI was founded in May 2014 through the collaboration of its co-founders, most prominently Tegmark and Tallinn. Tegmark — an MIT professor of physics whose research had turned from cosmology toward questions of the long-term future of intelligence in the universe — provided the academic credibility and research infrastructure. Tallinn — the Skype co-founder whose 2009 AI-doom realization had led him to become the principal funder of technical AI safety research worldwide — provided seed funding and strategic direction. The founding coincided with a broader intellectual moment in which AI safety was transitioning from a Rationalist-community concern to a mainstream technology-policy question. Stuart Russell's work on value alignment, Nick Bostrom's forthcoming *Superintelligence* book (published September 2014, three months after FLI's founding), and the beginning of large-scale deep learning breakthroughs had created appetite for a policy-adjacent organization that could translate technical AI safety concerns into advocacy. ## Advisory Board and Credibility Architecture FLI distinguished itself from earlier AI safety organizations (MIRI, FHI) through its deliberate assembly of a **high-profile advisory board** drawn from outside the core Rationalist ecosystem: - **Stephen Hawking** (before his 2018 death): The organization's highest-profile advisor; Hawking had publicly warned about AI risk starting around 2014 - **Elon Musk**: Co-signatory of FLI's earliest open letters on AI safety; Musk made a $10 million donation to FLI in 2015 for AI safety grants - **Stuart Russell** (UC Berkeley): The leading academic voice in AI alignment; FLI served as an institutional home for research that would inform his 2019 book *Human Compatible* - **Morgan Freeman, Alan Alda**: Celebrity advisory members who gave FLI a cultural presence outside scientific circles - **Yoshua Bengio, Francesca Rossi, Martin Rees**: Leading AI researchers and scientists who joined the advisory structure The advisory architecture was strategic. By pulling figures from outside the AI safety movement's core community, FLI could present its positions as scientific consensus rather than movement advocacy — a significant reputational asset in policy settings. ## The Asilomar AI Principles (2017) In January 2017 FLI convened the **Asilomar Conference on Beneficial AI** at the Asilomar Conference Grounds in California — explicitly modeled on the 1975 Asilomar Conference on Recombinant DNA that had established biosafety principles for genetic-engineering research. The 2017 conference produced the **Asilomar AI Principles**, a 23-point document signed by 1,200+ AI researchers and other signatories covering: - Research issues (shared benefits, research funding, science-policy link) - Ethics and values (human values, personal privacy, liberty and privacy, shared prosperity) - Long-term issues (capability caution, importance of superintelligence, common good) The Asilomar Principles became the most-cited public statement of AI safety values for years after their publication, reproduced in policy documents, AI company codes of conduct, and academic literature. ## The March 2023 Pause Letter The organization's most politically consequential single action was the **March 22, 2023 open letter calling for a six-month pause on "giant AI experiments" more powerful than GPT-4**. The letter was authored and hosted by FLI and was signed (on release) by: - **Elon Musk** — the most politically visible signatory - **Steve Wozniak** — Apple co-founder - **Yoshua Bengio** — AI pioneer and Turing Award winner - **Stuart Russell, Max Tegmark, Jaan Tallinn** — principal FLI figures - ~1,000+ other scientists, researchers, and technology figures at launch, growing to 30,000+ signatories over subsequent weeks The letter argued that contemporary AI systems were advancing faster than safety research could keep pace with and that a coordinated pause was necessary to allow alignment research to catch up. The letter was: - **Covered globally**: Front-page news in major world outlets; translated into dozens of languages - **Contested within the AI community**: Many leading AI researchers did *not* sign (including most of OpenAI's senior staff, most of DeepMind's senior staff, Anthropic leadership); critics including [[Timnit Gebru]] and [[Émile Torres]] argued the letter's premises were themselves ideological - **Influential on policy**: The letter accelerated a regulatory conversation already underway and is cited in the subsequent White House October 2023 Executive Order on AI and in EU AI Act debates Whether the letter actually produced a pause in frontier AI development (it did not) or primarily functioned as a media-forcing event (it did), its role in moving AI safety into mainstream political discourse is documentary. ## Grantmaking FLI has operated a grantmaking program for AI safety research since 2015, when Musk's $10 million donation seeded the first major round. Grantees have included: - Academic AI safety research groups at Stanford, Berkeley, MIT, Oxford, and other universities - Technical AI alignment projects outside the Rationalist-adjacent ecosystem - Policy-oriented AI safety research at think tanks and policy institutions - Biosecurity and nuclear-risk research adjacent to FLI's core agenda Total grantmaking is in the $10-20 million range cumulatively — smaller than [[Open Philanthropy]] or [[Jaan Tallinn]]'s SFF but meaningfully additive, particularly for academic groups that might not fit the funding priorities of the other major AI safety funders. ## Critical Commentary FLI has faced several categories of critical commentary: - **TESCREAL critique**: Torres and Gebru have named FLI as part of the broader TESCREAL institutional architecture; the pause letter was cited in their analysis as an example of the movement's preferred rhetorical framing (existential-risk-centered AI safety) crowding out other AI harm framings (algorithmic bias, labor displacement, surveillance) - **Musk involvement controversy**: Critics have noted that Musk's endorsement and funding coexist with xAI, Musk's own frontier AI company, whose stated goals are inconsistent with the "pause" framing FLI promoted - **Scientific-consensus framing**: The organization's ability to present AI safety concerns as scientific consensus via its advisory board has been praised by supporters and criticized by others as a PR strategy that obscures the movement's specific ideological lineage - **Russia/Ukraine controversy (2022)**: A small grant controversy related to Nya Dagbladet Foundation — a Swedish entity with connections the organization later publicly disavowed — generated press coverage that FLI handled through public apology and grant reversal ## Connection to the KB Ecosystem - **[[Longtermism]] / [[TESCREAL]]**: Principal advocacy organization for the AI-safety-framed longtermist agenda; the pause letter is one of the TESCREAL-aligned movement's most visible political actions. - **[[Effective Altruism]]**: Overlapping funder base and personnel; FLI is part of the EA-longtermist institutional ecosystem while maintaining broader external appeal. - **[[Max Tegmark]]**: Co-founder and principal public face of the organization. - **[[Jaan Tallinn]]**: Co-founder and principal funder; FLI is one of the organizations Tallinn's SFF supports. - **[[Elon Musk]]**: Early funder ($10M in 2015); pause letter signatory; subsequent xAI founder — the cross-cutting relationship between Musk and the AI safety ecosystem is partially mediated through FLI. - **[[MIRI]] / [[Future of Humanity Institute]]**: Adjacent institutions in the AI safety ecosystem; FLI has the most mainstream public presence of the three. - **[[Open Philanthropy]]**: Separate funder and grantmaker in the same space; FLI's scale is smaller but its public visibility is higher. - **AI pause debate**: The March 2023 pause letter is the landmark public-advocacy action; its effects on the political trajectory of AI regulation remain a live question through 2026. ## See Also - [[Max Tegmark]] - [[Jaan Tallinn]] - [[Elon Musk]] - [[Nick Bostrom]] - [[Stuart Russell]] - [[MIRI]] - [[Future of Humanity Institute]] - [[Centre for Effective Altruism]] - [[Open Philanthropy]] - [[Longtermism]] - [[Effective Altruism]] - [[TESCREAL]]