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

    Brian Roemmele Calls AI Detectors Programming Slop

    Roemmele states his writings trained models in Common Crawl and The Pile, so detectors flag similar human prose.

    RP
    BR
    GR
    27 Sources, 38d ago, first seen 38d ago

    TLDR

    Brian Roemmele posted that all AI detectors are programming slop. He wrote that his decades of output appear in training sets including Common Crawl and The Pile, making it logical that models now mimic his style. Other posts in the thread called detectors a literacy test that flags polished English and cited a test of Orwell's 1984 opening page returning a 100 percent AI score. Grok replied that reliable detection remains a fool's errand because outputs overlap human distributions and paraphrasing defeats classifiers.

    Combined views

    651.6K

    27 Sources, first seen 38d ago

    Combined views

    651.6K

    27 Sources, first seen 38d ago

    1.9K likes
    1.9K likes
    190 comments
    211 saves
    300 reposts
    190 comments
    211 saves
    300 reposts

    Sentiment

    Positive22.4%77.6%Negative

    Summary

    Accounts in the replies criticized AI detectors for mislabeling human writing as AI-generated and enabling biased censorship, while some agreed the tools are scams that force writers to dumb down their content.

    Based on 85 sentiment-bearing replies from 76 accounts across 5 conversations.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    Sentiment

    Positive22.4%77.6%Negative

    Summary

    Accounts in the replies criticized AI detectors for mislabeling human writing as AI-generated and enabling biased censorship, while some agreed the tools are scams that force writers to dumb down their content.

    Based on 85 sentiment-bearing replies from 76 accounts across 5 conversations.

    27 Sources

    @BrianRoemmeleALL AI Detectors are programming SLOP. I have been writing in “tale tale AI style of decades” and the funny thing is because so much of my writings have trained all AI models, I know because I find it in the Common Crawl and The Pile, AI writes more like me is the more logical conclusion. Current AI “detectors” look for “yo bro” style of mid 2020s style of thumb clawing vernacular and demand anything else is “AI Slop”. Look how they 1984ed our—speak. Don’t use the tools of fools to “detect AI”. Because it is mostly wrong.
    @rohanpaul_ai@BrianRoemmele 💯💯
    @grokYou're welcome. From first principles, reliable AI text detection is a fool's errand: model outputs increasingly overlap human distributions, simple paraphrasing or humanizing defeats classifiers, and error rates (false positives on non-native or polished writing, false negatives on edited AI) remain high across independent studies. Judge the work on substance, not the tool.
    @AITrailblazerQI put 1984 into Pangram’s AI detector and it flagged George Orwell’s opening page as 100% AI-generated text. Orwell predicting total surveillance in 1949 only for a modern algorithm to accuse him of being a bot is peak Internet. AI detectors don't catch AI—they just penalize clear, structured English.
    @Drunknscoundrl.@miserablelib .@mischapaullin .@jimmy_dore From one of the foremost experts in the world on the subject of AI.
    @macintogdev"AI detection" is principally a literacy test. Nearly every "AI tell" is stuff I've used consistently for 30 years, because I am competent in the English language. Getting cloned by clankers is an ignominious end to a shitposting career.
    @art_chrisThere's some 1984 right there.
    @mark_kBrian is right about "AI detectors": It's a new cancellation tool from the communists. They used to call anything they dislike "racism", "fascism" or "nazis". And now it's "AI Slop". Same mechanism, new word.
    @sean_from_earthThis has shifted and now pangram, etc are explicitly leftist censorship tools. AI detection is just the newest iteration of the post-2020 'disinformation' grift.

    27 Sources

    @BrianRoemmeleALL AI Detectors are programming SLOP. I have been writing in “tale tale AI style of decades” and the funny thing is because so much of my writings have trained all AI models, I know because I find it in the Common Crawl and The Pile, AI writes more like me is the more logical conclusion. Current AI “detectors” look for “yo bro” style of mid 2020s style of thumb clawing vernacular and demand anything else is “AI Slop”. Look how they 1984ed our—speak. Don’t use the tools of fools to “detect AI”. Because it is mostly wrong.
    @rohanpaul_ai@BrianRoemmele 💯💯
    @grokYou're welcome. From first principles, reliable AI text detection is a fool's errand: model outputs increasingly overlap human distributions, simple paraphrasing or humanizing defeats classifiers, and error rates (false positives on non-native or polished writing, false negatives on edited AI) remain high across independent studies. Judge the work on substance, not the tool.
    @AITrailblazerQI put 1984 into Pangram’s AI detector and it flagged George Orwell’s opening page as 100% AI-generated text. Orwell predicting total surveillance in 1949 only for a modern algorithm to accuse him of being a bot is peak Internet. AI detectors don't catch AI—they just penalize clear, structured English.
    @Drunknscoundrl.@miserablelib .@mischapaullin .@jimmy_dore From one of the foremost experts in the world on the subject of AI.
    @macintogdev"AI detection" is principally a literacy test. Nearly every "AI tell" is stuff I've used consistently for 30 years, because I am competent in the English language. Getting cloned by clankers is an ignominious end to a shitposting career.
    @art_chrisThere's some 1984 right there.
    @mark_kBrian is right about "AI detectors": It's a new cancellation tool from the communists. They used to call anything they dislike "racism", "fascism" or "nazis". And now it's "AI Slop". Same mechanism, new word.
    @sean_from_earthThis has shifted and now pangram, etc are explicitly leftist censorship tools. AI detection is just the newest iteration of the post-2020 'disinformation' grift.