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

    Andrew Carr Questions Need for AI Watermarks

    AI researcher highlights poor quality of recent frontier model text as possible detection aid.

    AC
    1 Source, 30d ago, first seen 30d ago

    TLDR

    Andrew Carr, co-founder and chief scientist at Cartwheel with past roles at OpenAI, posted on X about efforts to identify AI-generated writing through watermarks. He observed that many groups inject such markers into outputs. Referencing the degraded state of current frontier models, which he termed turbo fried, Carr wondered if watermarks are still required. The tweet stands alone without additional context or responses noted.

    Combined views

    1.3K

    1 Source, first seen 30d ago

    Combined views

    1.3K

    1 Source, first seen 30d ago

    18 likes
    18 likes
    2 comments
    2 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    2 comments
    2 reposts
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1 Source

    @andrew_n_carrThere has been discussion recently about identifying AI written text. And many groups are injecting watermarks into the model outputs to ensure that AI writing can be identified. Given the recent turbo fried writing of the frontier models,makes me wonder if we even need it?

    1 Source

    @andrew_n_carrThere has been discussion recently about identifying AI written text. And many groups are injecting watermarks into the model outputs to ensure that AI writing can be identified. Given the recent turbo fried writing of the frontier models,makes me wonder if we even need it?