Vigilia.
← Les avertissements
Risque d’extinctionLaboratoire de pointeMember of Technical Staff, Anthropic4 entrées au registre

Suivre

Holden Karnofsky

Cofondateur de GiveWell et d’Open Philanthropy, auteur de la série « Most Important Century » et, depuis 2025, membre de l’équipe technique d’Anthropic, il avertit que courir pour construire une IA transformatrice aussi vite que possible risque la mise à l’écart permanente de l’humanité.

Fonctions

  1. 2003 – 2007Analyst, Bridgewater AssociatesSource →
  2. 2007 – aujourd’huiCo-founder, GiveWellSource →
  3. 2014 – 2023Co-founder, CEO and later co-CEO, Open PhilanthropyTook leave in 2023, then moved to a Director of AI Strategy role.Source →
  4. 2017 – 2021Board member, OpenAISource →
  5. 2024 – janv. 2025Visiting Scholar, Carnegie Endowment for International PeaceWorked on international security risks from AI, including if-then commitments.Source →
  6. janv. 2025 – aujourd’huiMember of Technical Staff, AnthropicWorks on the Responsible Scaling Policy and AI safety planning.Source →

Au registre

  1. 30 oct. 2025 · Podcast · 80,000 Hours Podcast

    We're just racing to do this as fast as we possibly can. That seems like a really, really bad way to handle this technology.

    Member of Technical Staff, Anthropic à l’époque

    Recorded 25 and 28 July 2025, nine months into his Anthropic role.

    Holden Karnofsky on dozens of opportunities to make AI safer lying on the table — and all his AGI takes →

  2. 13 sept. 2024 · Article scientifique · Carnegie Endowment for International Peace

    AI could pose a variety of catastrophic risks to international security in several domains, including the proliferation and acceleration of cyberoffense capabilities

    Visiting Scholar, Carnegie Endowment for International Peace à l’époque

    If-Then Commitments for AI Risk Reduction →

  3. 9 juin 2022 · Publication · Cold Takes

    the kind of AI I've discussed could defeat all of humanity combined, if (for whatever reason) it were pointed toward that goal.

    Co-founder, CEO and later co-CEO, Open Philanthropy à l’époque

    AI Could Defeat All Of Us Combined →

  4. 13 juil. 2021 · Publication · Cold Takes

    this century could determine the entire future of the galaxy for tens of billions of years, or more.

    Board member, OpenAI à l’époque

    Opening post of the "Most Important Century" series.

    All Possible Views About Humanity's Future Are Wild →