• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    U.S.-China AI safety cooperation and concerns about chip exports

    In a September 17 post, a longtime supporter of U.S.-China AI safety cooperation worried an upcoming meeting could make the chip export situation worse.

    Miles BrundageMB
    Seán Ó hÉigeartaighSÓ
    Kristy LokeKL
    3 Sources, ,

    TLDR

    A user who says they have supported U.S.-China cooperation on AI safety for a decade-plus expressed strong skepticism about an upcoming meeting. Writing on September 17, they worried the meeting would only make the chip export situation worse, despite their broader support for cooperation.

    Combined views

    5.5K

    3 Sources, first seen 20d ago

    Combined views

    5.5K

    3 Sources, first seen 20d ago

    58 likes
    20d ago
    first seen 20d ago
    58 likes
    6 comments
    10 saves
    5 reposts
    6 comments
    10 saves
    5 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    Miles Brundage@Miles_BrundageI've been pro US-China cooperation on AI safety for a decade+ and written about it at length various times, but agree that one should be extremely skeptical of the upcoming meeting. I worry it will just lead to making the chip export situation even worse20d
    Kristy Loke@kristy_lokeLove your work, @EdwardGLuce and @AsiaLens, but I have a less pessimistic outlook. The fact that Washington and Beijing can still engage with each other on an issue so tightly bound up with national security, economic competition, and geopolitics is itself significant. It suggests a shared recognition that mismanagement and a lack of bilateral understanding/coordination carry real risks. Both sides already have enough regulatory capacity, technical expertise, as well as conceptual overlap to begin substantive discussions, e.g. around aligning on risk terminology, identifying near-horizon risks that may require communication or joint action, and gradually sharing emerging frontier model safety testing & hardening best practices. Emphasis on: enough & begin. One step at a time. As risks become increasingly felt on both sides (since China's moving towards the hidden global frontier, i.e. the capability of models that reside within U.S. frontier companies that they are not yet releasing), the two governments can potentially move quite quickly together. Just like Trump did post Mythos. Plus, I've been researching global AI governance for five years, and I can definitely say we've come a long way in a relatively short period of time! See my lengthier argument & proposal here:19d
    Seán Ó hÉigeartaigh@S_OhEigeartaighRT @kristy_loke: Love your work, @EdwardGLuce and @AsiaLens, but I have a less pessimistic outlook. The fact that Washington and Beijing c…19d

    3 Sources

    Miles Brundage@Miles_BrundageI've been pro US-China cooperation on AI safety for a decade+ and written about it at length various times, but agree that one should be extremely skeptical of the upcoming meeting. I worry it will just lead to making the chip export situation even worse20d
    Kristy Loke@kristy_lokeLove your work, @EdwardGLuce and @AsiaLens, but I have a less pessimistic outlook. The fact that Washington and Beijing can still engage with each other on an issue so tightly bound up with national security, economic competition, and geopolitics is itself significant. It suggests a shared recognition that mismanagement and a lack of bilateral understanding/coordination carry real risks. Both sides already have enough regulatory capacity, technical expertise, as well as conceptual overlap to begin substantive discussions, e.g. around aligning on risk terminology, identifying near-horizon risks that may require communication or joint action, and gradually sharing emerging frontier model safety testing & hardening best practices. Emphasis on: enough & begin. One step at a time. As risks become increasingly felt on both sides (since China's moving towards the hidden global frontier, i.e. the capability of models that reside within U.S. frontier companies that they are not yet releasing), the two governments can potentially move quite quickly together. Just like Trump did post Mythos. Plus, I've been researching global AI governance for five years, and I can definitely say we've come a long way in a relatively short period of time! See my lengthier argument & proposal here:19d
    Seán Ó hÉigeartaigh@S_OhEigeartaighRT @kristy_loke: Love your work, @EdwardGLuce and @AsiaLens, but I have a less pessimistic outlook. The fact that Washington and Beijing c…19d