Vigilia.
← Las advertencias
Riesgo de extinciónSociedad civilTechnical advisor, Center for AI Standards and Innovation, NIST3 entradas en el registro

Seguir

Paul Christiano

Investigador de alineamiento que dirigió el equipo de alineamiento de modelos de lenguaje de OpenAI, fundó el Alignment Research Center y desde 2024 dirigió la seguridad de la IA en el US AI Safety Institute; sitúa la probabilidad de una toma de control por la IA en decenas de puntos porcentuales.

Cargos

  1. 2017 – 2021Researcher; led the language model alignment team, OpenAICo-authored foundational work on reinforcement learning from human feedback.Fuente →
  2. abr 2021 – actualidadFounder and Executive Director, Alignment Research CenterCurrently listed as executive director on his own site, after returning from government.Fuente →
  3. sept 2023 – actualidadAdvisory board member, UK Frontier AI TaskforceFuente →
  4. 16 abr 2024 – 2026Head of AI Safety, US AI Safety Institute, NIST (renamed Center for AI Standards and Innovation, June 2025)Reported in August 2026 to have stepped back to a part-time advisor role in July 2026; the reporting (X) could not be fetched, but his site now lists him as technical advisor.Fuente →
  5. 2026 – actualidadTechnical advisor, Center for AI Standards and Innovation, NISTFuente →

En el registro

  1. 24 oct 2023 · Publicación · LessWrong

    I think AI developers are not prepared to work with very powerful AI systems. They don't have the scientific understanding to deploy superhuman AI systems without considerable risk

    Advisory board member, UK Frontier AI Taskforce en aquel momento

    Argues that a good responsible scaling policy must name the conditions under which development would be paused.

    Thoughts on responsible scaling policies and regulation →

  2. 27 abr 2023 · Publicación · LessWrong

    Probability of an AI takeover: 22% … Probability that humanity has somehow irreversibly messed up our future within 10 years of building powerful AI: 46%

    Founder and Executive Director, Alignment Research Center en aquel momento

    Also gives a 20% probability that most humans die within 10 years of building powerful AI.

    My views on "doom" →

  3. 19 jun 2022 · Publicación · LessWrong

    Powerful AI systems have a good chance of deliberately and irreversibly disempowering humanity. This is a much more likely failure mode than humanity killing ourselves with destructive physical technologies.

    Founder and Executive Director, Alignment Research Center en aquel momento

    First of his listed points of agreement with Yudkowsky.

    Where I agree and disagree with Eliezer →