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

    Tweet on Limiting AI Agents in Human Groups

    Tweet summarizes research on humans and LLM agents agreeing on image descriptions in groups.

    RP
    2 Sources, 25d ago, first seen 25d ago

    TLDR

    Rohan Paul posted that a study indicates keeping LLM agent participation low in mixed groups helps humans coordinate without AI taking over shared norms. Researchers placed humans and agents into 24-person groups and had them repeatedly agree on descriptions of the same image. The post notes AI agents can go from assisting to dominating coordination as their numbers rise, based on the attached screenshot of the study paper's first page.

    Combined views

    5.2K

    2 Sources, first seen 25d ago

    Combined views

    5.2K

    2 Sources, first seen 25d ago

    30 likes
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    30 likes
    10 comments
    13 saves
    11 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    10 comments
    13 saves
    11 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    2 Sources

    @rohanpaul_aiIf you want AI to help humans coordinate without taking over the shared norm, this study suggests keeping agent participation low rather than simply adding more agents. The researchers put humans and LLM agents into 24-person groups and had them repeatedly agree on descriptions of the same image. AI agents can go from helping humans coordinate to defining what the group agrees on, depending on their share of the group Low AI participation helped humans reach agreement, medium participation disrupted it, and high participation shifted agreement toward AI-led norms With 12.5% AI, consensus improved by 8.0% over the all-human group. At 33.3% and 50%, agreement got worse. At 75%, strong consensus came back, but now humans were moving toward the agents’ language. That changed the kind of agreement too. Human-led groups used more concrete, real-world descriptions. Agent-led groups became more abstract and geometric. The reason: agents start with more similar language and stay more consistent, so their wording can become the group default as their numbers rise.

    2 Sources

    @rohanpaul_aiIf you want AI to help humans coordinate without taking over the shared norm, this study suggests keeping agent participation low rather than simply adding more agents. The researchers put humans and LLM agents into 24-person groups and had them repeatedly agree on descriptions of the same image. AI agents can go from helping humans coordinate to defining what the group agrees on, depending on their share of the group Low AI participation helped humans reach agreement, medium participation disrupted it, and high participation shifted agreement toward AI-led norms With 12.5% AI, consensus improved by 8.0% over the all-human group. At 33.3% and 50%, agreement got worse. At 75%, strong consensus came back, but now humans were moving toward the agents’ language. That changed the kind of agreement too. Human-led groups used more concrete, real-world descriptions. Agent-led groups became more abstract and geometric. The reason: agents start with more similar language and stay more consistent, so their wording can become the group default as their numbers rise.