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

    Daniel Kokotajlo reportedly warns against trusting CEOs with superintelligent AI

    Andrew Curran’s account of hearing testimony raises concerns about both losing control and concentrating power.

    GM
    AC
    4 Sources, 2h ago, first seen 2h ago

    TLDR

    Andrew Curran quotes Daniel Kokotajlo as warning that tech CEOs probably could not control future, recursively self-improving superintelligent AI. Even if they retained control, Kokotajlo reportedly argued, they should not be trusted with it because they could become dictators or oligarchs in the United States. The warning describes a hypothetical risk; the quotation comes from Curran’s account of testimony.

    Combined views

    7.9K

    4 Sources, first seen 2h ago

    Combined views

    7.9K

    4 Sources, first seen 2h ago

    55 likes

    Useful links

    Business Insider

    A former OpenAI employee says the rush to beat China could push unsafe AI into the world

    Business Insider · YouTube

    Former OpenAI Researcher Warns 'AI Is Not Loyal To Us' | AI Architects | Business Insider

    Business Insider

    Former OpenAI researcher warns 'AI is not loyal to us'

    80,000 Hours · YouTube

    An AI 2027 Author Returns, With a Plan to Change the Ending
    Today's Rank

    #18

    Today's Rank

    #18

    55 likes
    5 comments
    5 saves
    6 reposts
    5 comments
    5 saves
    6 reposts

    5 Sources

    YouTubeLIVE: Former Anthropic researcher testifies on AI at New York City Council5h
    @AndrewCurran_Sworn testimony from Daniel Kokotajlo; 'Even if hypothetically these tech CEO's managed to maintain control of their superintelligent AI's that are recursively self improving, which again I think is a very big if - almost certainly they will not be able to maintain control - even if they 𝘥𝘰 manage to maintain control, I don't think we should trust them with that control. I think that there's a very real chance that they could use it to become dictators or oligarchs here in the United States.'2h
    @GaryMarcusOoh!!! @SpeakerMenin asks the tech firms to testify under oath that their systems will comply with their safety instructions. And they can’t! Why? Because the tech is a sloppy mess.1h

    Daniel Kokotajlo reportedly warned that keeping future superintelligent AI under control would still leave a question about who should wield that power.

    Featured Source

    In an Oct. 5 account of testimony, Andrew Curran quotes Kokotajlo as doubting that tech CEOs could maintain control of recursively self-improving AI. In the quoted warning, the systems would improve themselves, and Kokotajlo puts executives’ chances of retaining control very low.

    His second concern applies even if they succeed. Curran quotes him warning of a “very real chance” that those executives could use that control to become dictators or oligarchs in the United States.

    An Associated Press livestream identifies the event as a New York City Council hearing. The quotation is from Curran’s account of testimony. It describes Kokotajlo’s view of a hypothetical risk. It gives no timeline for that scenario.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Useful Links

    Business Insider

    A former OpenAI employee says the rush to beat China could push unsafe AI into the world

    Business Insider

    Former OpenAI researcher warns 'AI is not loyal to us'

    Related Videos

    • Former OpenAI Researcher Warns 'AI Is Not Loyal To Us' | AI Architects | Business InsiderBusiness Insider · YouTube
    • An AI 2027 Author Returns, With a Plan to Change the Ending80,000 Hours · YouTube

    5 Sources

    YouTubeLIVE: Former Anthropic researcher testifies on AI at New York City Council5h
    @AndrewCurran_Sworn testimony from Daniel Kokotajlo; 'Even if hypothetically these tech CEO's managed to maintain control of their superintelligent AI's that are recursively self improving, which again I think is a very big if - almost certainly they will not be able to maintain control - even if they 𝘥𝘰 manage to maintain control, I don't think we should trust them with that control. I think that there's a very real chance that they could use it to become dictators or oligarchs here in the United States.'2h
    @GaryMarcusOoh!!! @SpeakerMenin asks the tech firms to testify under oath that their systems will comply with their safety instructions. And they can’t! Why? Because the tech is a sloppy mess.1h

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Useful Links

    Business Insider

    A former OpenAI employee says the rush to beat China could push unsafe AI into the world

    Related Videos

    Business Insider

    Former OpenAI researcher warns 'AI is not loyal to us'
    Former OpenAI Researcher Warns 'AI Is Not Loyal To Us' | AI Architects | Business InsiderBusiness Insider · YouTube
  • An AI 2027 Author Returns, With a Plan to Change the Ending80,000 Hours · YouTube