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

    The AI-risk debate beyond human extinction

    A post describes Gary Marcus’s concern that flawed AI could cause catastrophes or erode privacy and democratic institutions.

    Gary MarcusGM
    Mario NawfalMN
    2 Sources, ,

    TLDR

    A post says Gary Marcus puts the probability of AI-driven human extinction in the next several years at “very close to zero,” but worries about catastrophe and dystopia. His scenarios include hallucinated intelligence influencing military decisions and cheap AI agents spreading propaganda. He also warns that AI could erode privacy, reliable information and democratic institutions.

    Combined views

    121.3K

    2 Sources, first seen 2h ago

    Combined views

    121.3K

    2 Sources, first seen 2h ago

    271 likes
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    2h ago
    first seen 2h ago
    271 likes
    31 comments
    31 saves
    45 reposts
    31 comments
    31 saves
    45 reposts

    2 Sources

    Mario Nawfal@MarioNawfal🇺🇸 AI probably won’t wipe out humanity. The real danger may be much closer. Gary Marcus thinks the AI debate is obsessing over the wrong nightmare. He puts the probability of AI-driven human extinction in the next several years at “very close to zero.” But that does not make him relaxed about AI. Marcus separates the risk into three categories: P-Doom, extinction; P-Catastrophe, events capable of killing huge numbers of people; and P-Dystopia, the gradual destruction of privacy, reliable information and democratic institutions. “Don’t panic about P-Doom, but worry a lot about P-Catastrophe and P-Dystopia.” The catastrophe scenario is disturbingly mundane. An AI hallucination contaminates an intelligence report. A military acts on false information. An automated system disrupts an electrical grid. A malicious actor uses cheap AI agents to manufacture propaganda or trigger a crisis. Marcus says an AI-generated intelligence error has already brought the world uncomfortably close to a military confrontation. “We’re rolling the dice here as a species.” And then there is dystopia. Marcus worries AI companies are accumulating enormous quantities of private information while governments and corporations gain increasingly powerful tools for shaping what people see and believe. His argument is almost the reverse of the usual AI apocalypse story. The machines do not need to become superintelligent and exterminate us. AI can remain flawed, unreliable and very human-controlled, and still become extraordinarily dangerous. @GaryMarcus2h
    Gary Marcus@GaryMarcusRT @MarioNawfal: 🇺🇸 AI probably won’t wipe out humanity. The real danger may be much closer. Gary Marcus thinks the AI debate is obsessing…1h
    Featured Source

    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.

    2 Sources

    Mario Nawfal@MarioNawfal🇺🇸 AI probably won’t wipe out humanity. The real danger may be much closer. Gary Marcus thinks the AI debate is obsessing over the wrong nightmare. He puts the probability of AI-driven human extinction in the next several years at “very close to zero.” But that does not make him relaxed about AI. Marcus separates the risk into three categories: P-Doom, extinction; P-Catastrophe, events capable of killing huge numbers of people; and P-Dystopia, the gradual destruction of privacy, reliable information and democratic institutions. “Don’t panic about P-Doom, but worry a lot about P-Catastrophe and P-Dystopia.” The catastrophe scenario is disturbingly mundane. An AI hallucination contaminates an intelligence report. A military acts on false information. An automated system disrupts an electrical grid. A malicious actor uses cheap AI agents to manufacture propaganda or trigger a crisis. Marcus says an AI-generated intelligence error has already brought the world uncomfortably close to a military confrontation. “We’re rolling the dice here as a species.” And then there is dystopia. Marcus worries AI companies are accumulating enormous quantities of private information while governments and corporations gain increasingly powerful tools for shaping what people see and believe. His argument is almost the reverse of the usual AI apocalypse story. The machines do not need to become superintelligent and exterminate us. AI can remain flawed, unreliable and very human-controlled, and still become extraordinarily dangerous. @GaryMarcus2h
    Gary Marcus@GaryMarcusRT @MarioNawfal: 🇺🇸 AI probably won’t wipe out humanity. The real danger may be much closer. Gary Marcus thinks the AI debate is obsessing…1h