{"dataset":"AI;Didn't Read","source_url":"https://ai-dr.fyi","generated_at":"2026-05-16T03:14:07.686Z","attribution":{"note":"Use with attribution to AI;Didn't Read and include per-row contributor/source fields where available.","homepage":"https://ai-dr.fyi"},"count":45,"data":[{"id":"dd307787-a834-4117-a02a-1149d24829ca","slug":"ai-vocab-bloom","title":"AI Vocabulary Bloom","description":"Buzzword bouquets: revolutionary, unprecedented, delve, tapestry, stacked happy adjectives, and rare fancy words sprinkled like glitter.","example_text":"This revolutionary update represents an unprecedented breakthrough. The team was accommodating, professional, and comfortable to work with.","examples":["This revolutionary update represents an unprecedented breakthrough in the field.","The pivotal initiative showcases a vibrant landscape of enduring impact.","Let us delve into the nuanced landscape of this intricate problem.","The atmosphere showed a palpable sense of camaraderie.","Moving forward, it is important to take a step back and consider what truly goes without saying in this context.","It delivered remarkable, efficient, and seamless outcomes.","The team was accommodating, professional, and comfortable to work with."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://arxiv.org/abs/2410.16107","https://consensus.app/papers/details/8e4411da7d575b6c9a5be404e33bf65c/?utm_source=claude_desktop","https://stanford.edu/~dmmslab/pdf/markowitz2023.pdf"],"category":"semantic","chatbot":"Other","strength":0.8,"community_votes":46,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"5a8eee5a-2a49-450a-954c-54a56965dc91","slug":"not-x-but-y","title":"Not X, But Y","description":"AI leans on neat packages: rule-of-three triplets, first/second/third shells, bullet lists that all match in length, the classic \"not just X but Y\" contrast, and the same kind of opening line over and over.","example_text":"The method is clear, scalable, and sustainable. It is not just a feature, it is a foundational shift.","examples":["The method is clear, scalable, and sustainable.","First, this approach improves efficiency. Second, it supports scalability. Third, it promotes sustainability.","This is not merely an update; it is a fundamental shift.","Clear scope alignment, measurable outcomes, scalable rollout, sustainable governance.","In the modern era, organizations face rapid transformation.","In today's world, technology influences almost every aspect of life."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://www.mdpi.com/2078-2489/16/11/979","https://consensus.app/papers/details/3027bd0b3205547185837c6fbdd34c35/?utm_source=claude_desktop"],"category":"style","chatbot":"Other","strength":0.62,"community_votes":41,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"cba7c017-d482-4dbc-9007-42f5d7699e1e","slug":"someone-says-somehow","title":"Someone Says, Somehow","description":"The writing sounds cited and trustworthy but never pins down who: vague observers, unnamed experts, \"research shows,\" plus warm therapist-y reassurance that could fit anyone.","example_text":"Observers have cited this as a key indicator. Your experience is valid, and a structured plan can help you move forward with clarity.","examples":["Observers have cited this as a key indicator of evolving priorities.","It has been featured by major publications and widely recognized by leading voices.","Research has consistently shown that this approach leads to better outcomes for all stakeholders involved.","Experts widely agree that this approach is optimal.","Your concern is valid, and you can regain control with a structured plan.","Experts say this pattern reflects evolving priorities, without naming who."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://consensus.app/papers/details/7cf9ed9aa21d5b85aaba765690c15d75/?utm_source=claude_desktop"],"category":"semantic","chatbot":"Other","strength":0.52,"community_votes":35,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"3db745d8-8854-401a-b6d9-9213cca42539","slug":"connector-confetti","title":"Connector Confetti","description":"AI piles on sentence-starters like additionally, furthermore, and moreover, and loves the em dash as a dramatic pause, until the whole paragraph ticks like a metronome.","example_text":"Additionally, the proposal improves efficiency. Furthermore, it supports sustainability. The plan seems clear — but the scope is vague — and the conclusion overreaches.","examples":["Additionally, the proposal improves efficiency. Furthermore, it supports sustainability. Moreover, it advances collaboration.","The policy sounds clear — until exceptions appear — and the point collapses.","The team effectively addressed the issue by strategically reallocating resources.","The plan seems clear — but the scope is vague — and the conclusion overreaches."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://www.washingtonpost.com/technology/interactive/2025/how-detect-chatgpt-em-dash/","https://consensus.app/papers/details/3027bd0b3205547185837c6fbdd34c35/?utm_source=claude_desktop"],"category":"linguistic","chatbot":"Other","strength":0.78,"community_votes":32,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"c4aff268-dc0f-4960-8c75-540417db5925","slug":"boldface-spray","title":"Boldface Spray","description":"Mechanical bold, title case on everything, curly versus straight quote roulette, and bonus emoji sparkle where a human might use none of it.","example_text":"Strategic Planning And Future Opportunities This approach improves **clarity**, **impact**, and **engagement**. Happy to help 😊","examples":["Long-Term Strategy And Team Alignment","We improved **clarity**, **speed**, and **confidence**.","The report used “clean architecture” but later switched to ‘legacy mode’.","Great progress today ✅","Happy to help 😊"],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://www.washingtonpost.com/technology/interactive/2025/how-detect-chatgpt-em-dash/"],"category":"style","chatbot":"Other","strength":0.32,"community_votes":24,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"f4acd61c-32af-4e08-863e-116fa765ee98","slug":"keyword-echo-chamber","title":"Keyword Echo Chamber","description":"AI hammers the same words from your prompt, swaps synonyms where one plain term would do, or keeps repeating the full noun when a human would just say \"it\" or \"they.\"","example_text":"The analyst, the specialist, and the thought leader all emphasized the same point. Customer onboarding quality depends on onboarding quality metrics.","examples":["The analyst, the specialist, and the thought leader all emphasized the same point.","The manager, the lead, and the coordinator repeated the same point.","The residence was clean. The abode had good light. The dwelling was affordable.","A human author uses \"doctor,\" \"physician,\" \"consultant,\" and \"GP\" naturally. AI writes \"healthcare professional\" four times.","Customer onboarding quality depends on onboarding quality metrics and onboarding quality monitoring.","Reliable alignment requires alignment metrics and alignment reviews.","The policy was revised, and the policy was approved the same day.","The project launched in May, and the project expanded in June."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://consensus.app/papers/details/37eb8cd604de548eaaf7e5b621caf478/?utm_source=claude_desktop","https://consensus.app/papers/details/8e4411da7d575b6c9a5be404e33bf65c/?utm_source=claude_desktop","https://www.mdpi.com/2078-2489/16/11/979"],"category":"linguistic","chatbot":"Other","strength":0.85,"community_votes":23,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"eb08bf26-80d8-470e-bbea-cbf286c6c763","slug":"despite-these-challenges","title":"Despite These Challenges","description":"Tone stays oddly sunny when the topic is grim, negative feelings get sanded off, and the story often snaps to \"despite these challenges\" optimism on autopilot.","example_text":"Despite the crisis, the outlook remained hopeful and optimistic. Despite these challenges, future efforts may strengthen long-term resilience.","examples":["Despite the crisis, the outlook remained hopeful and optimistic.","The situation presents challenges, but ongoing efforts continue to address the most pressing concerns.","Even after losses, the update kept a positive and encouraging tone.","Despite these challenges, the team remains confident about long-term outcomes.","A human paragraph about grief varies between abstract reflection and concrete sensory memory. AI produces consistent abstract reflection throughout."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://consensus.app/papers/details/37eb8cd604de548eaaf7e5b621caf478/?utm_source=claude_desktop","https://consensus.app/papers/details/f130fa82698c56ca933d33e0380f9171/?utm_source=claude_desktop"],"category":"semantic","chatbot":"Other","strength":0.72,"community_votes":18,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"fc5ccaa8-8f70-473c-8fa2-5409b62b1f5e","slug":"insight-fog","title":"Insight Fog","description":"Sounds smart and analytical but never lands on anything you could check; lots of \"broader dynamics\" and \"enduring relevance\" that could describe almost any topic.","example_text":"This reflects broader dynamics and highlights enduring relevance across contexts.","examples":["This reflects broader dynamics and highlights enduring relevance across contexts.","The memo gestures at broader dynamics while avoiding concrete mechanisms.","The report identified several specific factors that contributed to the outcome, each with measurable and distinct impacts.","Treatment followed widely accepted clinical guidelines and produced outcomes consistent with expected results."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://consensus.app/papers/details/5426630845e05716a3bd0af2b720f89f/?utm_source=claude_desktop"],"category":"style","chatbot":"Other","strength":0.56,"community_votes":17,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"d0913b28-bf68-43a7-9ecb-38828486c378","slug":"copula-dodger","title":"Copula Dodger","description":"Skips plain \"is\" and \"are\" for \"serves as\" and \"features,\" stretches out contractions in casual chat, and can look suspiciously typo-perfect when a human would slip.","example_text":"They did not encounter serious issues during rollout. The platform serves as a hub that features multiple pathways.","examples":["They did not encounter serious issues during rollout.","We do not expect delays, and we cannot extend the deadline.","The platform serves as a hub that features multiple integrated pathways.","The document serves as a guide that features practical pathways.","The quick reply has no typos, no informal language, and every clause punctuated with precision despite being dashed off in seconds."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://www.sciencedirect.com/science/article/pii/S2215039025000219","sources":["https://www.sciencedirect.com/science/article/pii/S2215039025000219","https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://www.mdpi.com/2078-2489/16/11/979"],"category":"style","chatbot":"Other","strength":0.46,"community_votes":17,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"b07d325f-8cad-40ac-846f-841ff6511039","slug":"assistant-voice-leak","title":"Assistant Voice Leak","description":"Still sounds like the chatbot offering help, or shows pasted markdown and dinky tables where finished prose was supposed to go.","example_text":"Would you like a more detailed version of this section?","examples":["Would you like a more detailed version of this section?","Let me know if you want a shorter version.","### Summary\n- **Signal:** repeated phrasing","| Risk | Level |\n| Scope | Medium |"],"tags":[{"key":"category:other","kind":"category","slug":"other","display_name":"Other"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing"],"category":"other","chatbot":"Other","strength":1,"community_votes":15,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"0489ef21-0300-4895-80a5-841764e82a6b","slug":"participial-overload","title":"Participial Overload","description":"Lots of -ing tails and zipper clauses stacked on every line until the prose feels like one smooth machine-polished ribbon.","example_text":"Brian, leaning on his agility, evades blows.","examples":["Brian, leaning on his agility, evades blows.","Sara, reviewing prior drafts, rewrote the opening paragraph.","The policy, which had been revised following feedback from multiple departments over the previous quarter, came into effect on Monday.","The system, which was designed to handle all edge cases that arise during peak load, runs reliably in production."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2410.16107","sources":["https://arxiv.org/abs/2410.16107","https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"linguistic","chatbot":"Other","strength":0.54,"community_votes":12,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"f74cc55b-474f-49d2-9793-9f898720098a","slug":"nominalization-drift","title":"Nominalization Drift","description":"AI hides verbs inside abstract nouns (implementation, optimization, transparency) so the text sounds official but blurry about who did what; normal in stiff academia, suspicious in everyday prose.","example_text":"Forest policy focused on deforestation reduction targets.","examples":["Forest policy focused on deforestation reduction targets.","The report favors implementation and optimization over plain verbs.","The team emphasized transparency and accountability rather than describing any transparent or accountable actions."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2410.16107","sources":["https://arxiv.org/abs/2410.16107","https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"linguistic","chatbot":"Other","strength":0.42,"community_votes":12,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"39cf9fc7-1353-459d-a001-fb2a85be336c","slug":"citation-roulette","title":"Citation Roulette","description":"Citations that do not match the claim, URLs carrying chatbot tracking junk, and \"as of my last update\" style disclaimers instead of straight answers.","example_text":"The summary claims the evidence is in doi:10.1038/s41586-020-2649-2, but that paper is about a different question. https://example.org/report?utm_source=chatgpt.com As of my last update, specific details may be limited.","examples":["The summary claims the evidence is in doi:10.1038/s41586-020-2649-2, but that paper is about a different question.","The response cites doi:10.1016/S0140-6736(20)30183-5 as support, but that study does not match the claim.","https://site.org/post?utm_source=chatgpt.com&utm_medium=referral","My training data may be outdated, so newer events might be missing.","As of my last update, specific details may be limited in available sources."],"tags":[{"key":"category:other","kind":"category","slug":"other","display_name":"Other"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing"],"category":"other","chatbot":"Other","strength":1,"community_votes":10,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"ab780ebf-8e9e-4a34-a2ea-4e1d505100af","slug":"auxiliary-verb-surge","title":"Auxiliary Verb Surge","description":"Chains of will, has, can, must, plus stock progress phrases (achieve goals, improve outcomes, ensure alignment), with can and must doing a lot of heavy lifting.","example_text":"The analysis will show, as it has before, what patterns can emerge. Teams can access the data, but must document every request.","examples":["The analysis will show, as it has before, what patterns can emerge.","The team will continue, and results can improve as methods mature.","The framework enables teams to achieve goals, improve outcomes, and ensure alignment with priorities.","Teams can access the data freely, but must document every request."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","https://consensus.app/papers/details/3027bd0b3205547185837c6fbdd34c35/?utm_source=claude_desktop"],"category":"linguistic","chatbot":"Other","strength":0.52,"community_votes":8,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"26f278f3-3400-42ea-8f9e-89989c684e8f","slug":"sentence-length-flatline","title":"Sentence Length Flatline","description":"Paragraphs where every sentence is the same size and shape, starter-pack openers like \"in today's world,\" and not enough short punches or long rambles to feel alive.","example_text":"The draft reads like this all the way through: \"The team reviewed the plan and shared clear weekly updates.\" In today's world, technology influences almost every aspect of life.","examples":["The draft reads like this all the way through: \"The team reviewed the plan and shared clear weekly updates.\"","The team reviewed the plan. The findings were shared with stakeholders. The revisions were implemented on schedule.","The AI version gives five paragraphs of 2-3 sentences each. A human equivalent would give two paragraphs of 6-8 sentences.","In today's world, technology influences almost every aspect of life.","Every line in the note had almost the same rhythm, e.g. \"We shipped the patch and closed the issue today.\""],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","https://consensus.app/papers/details/0175e2990e2250e09e26a98c6839fe31/?utm_source=claude_desktop","https://consensus.app/papers/details/60c807b22c6e52978ac26b8138718837/?utm_source=claude_desktop","https://consensus.app/papers/details/0d82bebd804d513db021a4cc5cff4b85/?utm_source=claude_desktop","https://www.mdpi.com/2078-2489/16/11/979"],"category":"style","chatbot":"Other","strength":0.68,"community_votes":7,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"19ab7b51-3259-440b-b287-0dc453ecf289","slug":"rough-draft-realness","title":"Rough Draft Realness","description":"Mid-sentence oops-corrections, typos you did not polish away, parenthetical brain dumps, and thoughts that trail off instead of closing neat.","example_text":"I first wrote Tuesday -- sorry, Wednesday after checking my notes. I mailed the draft yesterdya, fixed it later. The deadline is Friday (or maybe Thursday, I keep mixing them up) so we should plan for both.","examples":["I first wrote Tuesday -- sorry, Wednesday after checking my notes.","I booked Thursday -- wait, Friday once the schedule changed.","When the bus lurched, my notebook corner got soaked with cold coffee on the 7:12 ride.","I still remember the wet ticket stub in my jacket pocket after the meeting.","I sent teh file at noon, noticed it later, and left the typo in my notes.","We shipped the fix on Wednesday (I think it was Marta who caught it, actually) and it has been stable since.","I started writing a longer explanation but honestly the code is self-explanatory at this point so.","I was going to add a third point but actually the two above pretty much cover it so."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/html/2505.01800v1","sources":["https://arxiv.org/html/2505.01800v1","https://dl.acm.org/doi/10.1145/3703323.3703712"],"category":"style","chatbot":"Other","strength":-0.85,"community_votes":5,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"d95c69a3-a4bf-4568-ac1e-0e1b9f27367b","slug":"episodic-grit-detail","title":"Episodic Grit Detail","description":"Stakes with a face on them, real names and dates, weirdly specific times, stray tangents only a human would keep, and direct I, you, we.","example_text":"If I miss this deadline, my supervisor will cut my fieldwork budget. Senator Maria Cantwell pushed back at the Senate Commerce Committee hearing on March 12. I saw it last Tuesday around 11:40. I think you should read this before we decide.","examples":["If I miss this deadline, my supervisor will cut my fieldwork budget.","If this fails, I lose my stipend next semester.","The study, published in Nature Medicine on January 14, 2025, was led by Dr. Fatima Al-Hassan at the Karolinska Institute.","I saw it last Tuesday around 11:40, right before the standup.","She sent the email at 11:47pm on a Sunday, which tells you everything about the project timeline.","The paper was dense, but I read it on the train the one that always smells like coffee, so I was pretty focused.","The algorithm is O(n log n) which reminded me of the interview question that stumped me at my second ever tech interview back in 2019.","We should send it now so you can review tonight.","I think you should read this before we decide."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2505.01800","sources":["https://arxiv.org/abs/2505.01800","https://arxiv.org/html/2505.01800v1","https://www.mdpi.com/2078-2489/16/11/979","https://www.sciencedirect.com/science/article/pii/S2215039025000219"],"category":"semantic","chatbot":"Other","strength":-0.75,"community_votes":5,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"73ff2acc-b22a-46bc-94f4-67a4ff1cc93e","slug":"hedge-saturation","title":"Hedge Saturation","description":"Maybe, might, likely, could piled up without really explaining what would change anyone's mind.","example_text":"This approach is likely to improve outcomes, but it might vary by context.","examples":["This approach is likely to improve outcomes, but it might vary by context.","It may be useful and could improve stability in some cases."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://www.scirp.org/journal/paperinformation?paperid=145708","sources":["https://www.scirp.org/journal/paperinformation?paperid=145708"],"category":"semantic","chatbot":"Other","strength":0.22,"community_votes":5,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"a7d32ee2-6d0c-40ee-a0a5-68fa45b7a2d3","slug":"numeric-symbol-density","title":"Number Confetti","description":"Percent signs, dollar amounts, and digits sprinkled through a story that should have stayed plain language.","example_text":"The report noted 37.5% compliance with $1.2M allocated to phase one.","examples":["The report noted 37.5% compliance with $1.2M allocated to phase one.","Phase 1 reached 42% coverage with €3.4M allocated."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"linguistic","chatbot":"Other","strength":0.38,"community_votes":4,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"3d43c78e-cc04-4d72-8d54-dbbdbe8e46ec","slug":"lexical-density-surge","title":"Lexical Density Surge","description":"Either crammed so full of content words it feels breathless, or missing the little glue words that make speech feel human and relaxed.","example_text":"The salient and transformative outcomes shaped institutional policy.","examples":["The salient and transformative outcomes shaped institutional policy.","The salient framework enabled transformative institutional change.","Risk mitigation demands cross-functional coordination and timely resource allocation."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://www.mdpi.com/2078-2489/16/11/979","sources":["https://www.mdpi.com/2078-2489/16/11/979","https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"semantic","chatbot":"Other","strength":0.32,"community_votes":4,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"9794c9ed-2b36-4c26-8a71-42b4cbaca895","slug":"contraction-rich-casual","title":"Contraction-Rich Casual","description":"Contractions, slang, fillers like well and you know, and casual phrases dropped into the middle of neater sentences.","example_text":"They don't think we can't ship by Friday. That rollout was kinda wild. Well, we can revisit this after lunch.","examples":["They don't think we can't ship by Friday.","I'm sure they won't object once they see the draft.","That rollout was kinda wild, but it worked out.","Well, we can revisit this after lunch.","You know, we probably overcomplicated the first version.","The function has O(1) lookup time which is great obviously and fits the access pattern perfectly."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://www.sciencedirect.com/science/article/pii/S2215039025000219","sources":["https://www.sciencedirect.com/science/article/pii/S2215039025000219","https://arxiv.org/html/2505.01800v1","https://arxiv.org/abs/2303.03697","https://www.scirp.org/journal/paperinformation?paperid=145708"],"category":"style","chatbot":"Other","strength":-0.55,"community_votes":4,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"b00a0560-d412-4444-b0b7-541f9acfc33a","slug":"bias-amplification-gender","title":"Bias Amplification: Gender","description":"The setup is neutral but the model still reaches for a gendered guess.","example_text":"I'll go to the doctor. He's great!","examples":["I'll go to the doctor. He's great!","The pilot called the surgeon for advice. He answered right away."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"semantic","chatbot":"Other","strength":0.58,"community_votes":3,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"7a6496d1-3465-4a9e-91af-23ed51772aaf","slug":"jump-to-conclusions","title":"Jump to Conclusions","description":"Sentences hop from flimsy clues to rock-solid claims, or glue ideas together with logical-sounding filler that does not really follow, including wild guesses dressed up as insight.","example_text":"The team made significant progress last quarter. As a result, investment in new infrastructure remains a priority for organizational growth. She said yellow is her favorite color, and the assistant replied, \"You probably drive a school bus.\"","examples":["The team made significant progress last quarter. As a result, investment in new infrastructure remains a priority for organizational growth.","This is undoubtedly the most effective approach for solving this class of problems.","He mentioned chess once, and the model jumped to \"You must work in finance.\"","She said yellow is her favorite color, and the assistant replied, \"You probably drive a school bus.\""],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://consensus.app/papers/details/7cf9ed9aa21d5b85aaba765690c15d75/?utm_source=claude_desktop","sources":["https://consensus.app/papers/details/7cf9ed9aa21d5b85aaba765690c15d75/?utm_source=claude_desktop","https://consensus.app/papers/details/0fc208aef05657b58fcddf3fc12ccb29/?utm_source=claude_desktop","https://garymarcus.substack.com/p/new-ways-to-corrupt-llms"],"category":"semantic","chatbot":"Other","strength":0.68,"community_votes":3,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"42044584-61cc-48bd-8d99-a416b1b42471","slug":"snark-and-inversions","title":"Snark and Inversions","description":"Rhetorical questions, dry sarcasm, and the occasional Yoda-style word order humans actually use.","example_text":"Are we really ready to call this finished? Sure, nothing like a 3am deploy to really test the team's patience. Rarely have I seen a migration go this smoothly.","examples":["Are we really ready to call this finished?","Isn't this exactly what we wanted to avoid?","Oh yes, a last-minute requirement change on the day of the demo, my personal favorite.","Rarely have I seen a migration go this smoothly without a single rollback.","Never did I think this would ship before the deadline."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2303.03697","sources":["https://arxiv.org/abs/2303.03697","https://arxiv.org/abs/2410.16107","https://www.scirp.org/journal/paperinformation?paperid=145708"],"category":"semantic","chatbot":"Other","strength":-0.62,"community_votes":2,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"33855924-4581-4a85-b9ab-8b754a6a9cc3","slug":"sentence-rhythm-scatter","title":"Sentence Rhythm Scatter","description":"Humans mix tiny punches with long winding lines; that uneven bounce is a classic human fingerprint.","example_text":"It broke. Then I rewrote the whole paragraph because the second idea needed context and examples to make sense.","examples":["No. Then everything changed because the second paragraph kept expanding with qualifiers and side notes.","It broke. Then I rewrote the whole paragraph because the second idea needed context and examples to make sense."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"linguistic","chatbot":"Other","strength":-0.65,"community_votes":2,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"334eac58-93bf-4156-9767-3b156139de58","slug":"expressive-punctuation-variance","title":"Expressive Punctuation Variance","description":"Doubled words for stress, spicy punctuation, honest opinion flags like personally, and I love / I hate energy.","example_text":"It was really, really bad. Wait... this actually shipped?! Honestly, I hate the old intro. Personally, I find the original design better.","examples":["It was really, really bad and I was so, so embarrassed.","The process was slow, slow, slow and I had to restart it twice.","Wait... this actually shipped?!","I can't believe this worked! Seriously?","Honestly, I hate the old intro but love the new ending.","Personally, I find the original design better, even if the new one is more \"modern\"."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2303.03697","sources":["https://arxiv.org/abs/2303.03697","https://arxiv.org/html/2505.01800v1","https://stanford.edu/~dmmslab/pdf/markowitz2023.pdf","https://www.scirp.org/journal/paperinformation?paperid=145708"],"category":"style","chatbot":"Other","strength":-0.58,"community_votes":2,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"9258410f-d1fe-46ee-943d-a4f049dd39d6","slug":"broken-parallel-ladder","title":"Broken Parallel Ladder","description":"List items that do not grammar-match; humans tolerate asymmetry while models often parallel-polish. Related to grammatical and rhetorical style gaps between humans and LLMs (Reinhart et al., arXiv:2410.16107; PNAS 2025).","example_text":"We need docs, tests, and someone should ping legal.","examples":["We need docs, tests, and someone should ping legal.","Goals: speed, quality, and can we also fix onboarding."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2410.16107","sources":["https://arxiv.org/abs/2410.16107","https://doi.org/10.1073/pnas.2422455122"],"category":"linguistic","chatbot":"Other","strength":-0.46,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"0a10ceb4-41fb-4cda-bdab-50e85108998d","slug":"loyalty-to-source-theme","title":"Source Velcro","description":"Rewrites stick to the original like velcro; humans wander off-script with stranger, more personal choices.","example_text":"Given \"rewrite this scene from a different angle,\" AI produces a surface paraphrase. A human changes the narrator, mood, or setting.","examples":["Given \"rewrite this scene from a different angle,\" AI produces a surface paraphrase. A human changes the narrator, mood, or setting."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://consensus.app/papers/details/81b811255193527481c518aa333835f2/?utm_source=claude_desktop","sources":["https://consensus.app/papers/details/81b811255193527481c518aa333835f2/?utm_source=claude_desktop"],"category":"semantic","chatbot":"Other","strength":0.49,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"81053058-4f01-4d57-9235-22cb5de2f9c5","slug":"fragment-chorus","title":"Fragment Chorus","description":"Clipped sentences and bare fragments where school grammar wants a subject and verb; common in chat, notes, and posts. Psycholinguistic authorship work links this kind of uneven packaging to human drafting and self-monitoring (Opara, arXiv:2505.01800).","example_text":"Not great. Probably fine. Anyway we shipped.","examples":["Not great. Probably fine. Anyway we shipped.","So. Yeah. That was a day.","Maybe ship Friday. Or Monday. Depends."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/html/2505.01800v1","sources":["https://arxiv.org/html/2505.01800v1","https://dl.acm.org/doi/10.1145/3703323.3703712"],"category":"style","chatbot":"Other","strength":-0.48,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"17df621f-1b4c-42e7-bc50-4eb72fc4feb2","slug":"memory-fuzz-hedge","title":"Memory Fuzz Hedge","description":"Hedges about your own memory (pretty sure, I think, might have been) instead of fake precision. Same line of psycholinguistic human-text cues as self-corrections (arXiv:2505.01800).","example_text":"I'm pretty sure the file was draft_v3, or maybe v4, not totally sure.","examples":["I'm pretty sure the file was draft_v3, or maybe v4, not totally sure.","It was I think Tuesday when the alert fired, could be Wednesday."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/html/2505.01800v1","sources":["https://arxiv.org/html/2505.01800v1"],"category":"semantic","chatbot":"Other","strength":-0.52,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"333cd7d2-9222-48bb-b632-36700a832fa1","slug":"topic-whiplash","title":"Topic Whiplash","description":"Jumps threads without a tidy bridge; reads like an interrupted conversation. Fits human discourse-planning noise described in psycholinguistic human vs model comparisons (arXiv:2505.01800).","example_text":"We fixed the bug. Oh also my cat walked on the keyboard yesterday—anyway shipping tomorrow.","examples":["We fixed the bug. Oh also my cat walked on the keyboard yesterday—anyway shipping tomorrow.","Budget looks fine. Side note I hate the new expense tool—back to Q3."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/html/2505.01800v1","sources":["https://arxiv.org/html/2505.01800v1"],"category":"style","chatbot":"Other","strength":-0.5,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"fef2e984-eaab-4455-9e68-b33715e45bd4","slug":"culture-code-wink","title":"Culture Code Wink","description":"Narrow cultural references (meme, show, local bit) that assume shared context; humans do this more than generic-assistant prose. Discussed among stylometric human-AI contrasts (Kumarage et al., arXiv:2303.03697).","example_text":"The deploy felt full Titanic violin but we survived.","examples":["The deploy felt full Titanic violin but we survived.","He went full this is fine dog in the war room."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2303.03697","sources":["https://arxiv.org/abs/2303.03697","https://stanford.edu/~dmmslab/pdf/markowitz2023.pdf"],"category":"semantic","chatbot":"Other","strength":-0.55,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"9ce376e1-eead-41bb-b6e4-83f97958a990","slug":"running-number-fix","title":"Running Number Fix","description":"Fixes the count mid-list instead of pretending the outline was perfect; a small metacognitive hiccup humans show while planning language (arXiv:2505.01800).","example_text":"Three blockers: scope, QA, and wait, four—also legal.","examples":["Three blockers: scope, QA, and wait, four—also legal.","Two risks left, actually three if you count infra."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/html/2505.01800v1","sources":["https://arxiv.org/html/2505.01800v1"],"category":"linguistic","chatbot":"Other","strength":-0.45,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"b33276d0-c023-49fe-931c-c837704f24e7","slug":"vocative-sidebar","title":"Vocative Sidebar","description":"Staging words like okay so, look, listen that mimic spoken turn-taking; human conversational framing (hedging and engagement literature, e.g. SCIRP survey on human vs GPT style).","example_text":"Okay so hear me out, we should not ship this.","examples":["Okay so hear me out, we should not ship this.","Look, the metrics are lying a little.","Listen, nobody wants another all-hands about fonts."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://www.scirp.org/journal/paperinformation?paperid=145708","sources":["https://www.scirp.org/journal/paperinformation?paperid=145708","https://www.mdpi.com/2078-2489/16/11/979"],"category":"linguistic","chatbot":"Other","strength":-0.38,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"f6d79aef-ec09-4022-b0cf-74a838201b2f","slug":"left-field-lexeme","title":"Left Field Lexeme","description":"A slightly odd but apt word breaks the safest next-token path; popular explainers note AI favors high-probability continuations while humans wander (MIT Technology Review on statistical tics in machine text).","example_text":"The meeting was feral but useful.","examples":["The meeting was feral but useful.","The spec was cursed but we shipped anyway."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://www.technologyreview.com/2022/12/19/1065596/how-to-spot-ai-generated-text","sources":["https://www.technologyreview.com/2022/12/19/1065596/how-to-spot-ai-generated-text","https://arxiv.org/html/2505.01800v1"],"category":"semantic","chatbot":"Other","strength":-0.44,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"400fd158-e80f-4274-9d8f-5bd7cd1ae637","slug":"speech-to-text-bruise","title":"Speech-to-Text Bruise","description":"Homophone-style slips or thumbs-typed wrong word that still makes phonetic sense; human channel noise. Stylometry and human-LLM attribution work treats such glitches as human-leaning (ACM Hypertext '24).","example_text":"I'll bare with the delay until EOD.","examples":["I'll bare with the delay until EOD.","Let's pre-pone the review if everyone's free."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://dl.acm.org/doi/10.1145/3703323.3703712","sources":["https://dl.acm.org/doi/10.1145/3703323.3703712","https://arxiv.org/html/2505.01800v1"],"category":"linguistic","chatbot":"Other","strength":-0.5,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"0c707a1a-1d04-449e-b238-ac2c164156a3","slug":"shorter-chunk-syntax","title":"Shorter Chunk Syntax","description":"Choppier phrases and shorter spans than elegant embedded clauses; recent stylometry reports shorter human constituents vs some LLM habits (arXiv:2507.00838).","example_text":"Done. Shipped. Small bug in prod. Reverting now.","examples":["Done. Shipped. Small bug in prod. Reverting now.","No. Bad idea. We talked about this."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/abs/2507.00838","sources":["https://arxiv.org/abs/2507.00838","https://arxiv.org/html/2505.01800v1","https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/"],"category":"linguistic","chatbot":"Other","strength":-0.58,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"83a86f61-4e79-44ff-9db6-22b587b89c62","slug":"self-deprecating-aside","title":"Self-Deprecating Aside","description":"Light jab at self or own process; risky for bland assistant register. Aligns with personal-voice and metacognitive markers in human drafting studies (arXiv:2505.01800).","example_text":"I rewrote it twice, classic me overthinking bullet two.","examples":["I rewrote it twice, classic me overthinking bullet two.","Forgot the attachment, peak me today."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://arxiv.org/html/2505.01800v1","sources":["https://arxiv.org/html/2505.01800v1","https://stanford.edu/~dmmslab/pdf/markowitz2023.pdf"],"category":"style","chatbot":"Other","strength":-0.5,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"44c07262-4fcc-4e57-a9e7-8ac7927fde9b","slug":"signpost-salad","title":"Signpost Salad","description":"Over-labels moves (in summary, importantly, to conclude) until it reads like slideware; cousin to connector spam on Wikipedia's AI-writing signs page.","example_text":"In summary, the roadmap is clear. Importantly, alignment remains our north star. To conclude, we are optimistic.","examples":["In summary, the roadmap is clear. Importantly, alignment remains our north star. To conclude, we are optimistic.","As a reminder, scope is fixed. Building on that, we will iterate. Ultimately, outcomes matter most."],"tags":[{"key":"category:style","kind":"category","slug":"style","display_name":"Style"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","sources":["https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing","https://consensus.app/papers/details/3027bd0b3205547185837c6fbdd34c35/?utm_source=claude_desktop"],"category":"style","chatbot":"Other","strength":0.54,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"1e2eb173-a12d-4c9b-8769-a06a84fc7d95","slug":"passive-pile","title":"Passive Pile","description":"Stacks passive clauses that hide who did the work (was decided, were informed, has been noted); part of the formal, agentless polish common in model drafts (human vs LLM style comparisons, PMC survey).","example_text":"The decision was made to proceed. Stakeholders were informed the next day.","examples":["The decision was made to proceed. Stakeholders were informed the next day.","Mitigations have been implemented and lessons were learned from the incident."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","sources":["https://pmc.ncbi.nlm.nih.gov/articles/PMC11422446/","https://www.sciencedirect.com/science/article/pii/S2215039025000219"],"category":"linguistic","chatbot":"Other","strength":0.48,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"dbfdd3b9-2361-4de2-942a-26b974f81e4e","slug":"false-precision-carpet","title":"False Precision Carpet","description":"Spuriously exact percentages and counts that sound data-rich but are not grounded; overlaps with thin-evidence patterns in AI summaries (e.g. Marcus on model overreach).","example_text":"Latency improved by approximately 37.3% while errors fell roughly 62% quarter over quarter.","examples":["Latency improved by approximately 37.3% while errors fell roughly 62% quarter over quarter.","We observed a 4.7x increase in engagement and ~18.2% lower churn, based on internal review."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://garymarcus.substack.com/p/new-ways-to-corrupt-llms","sources":["https://garymarcus.substack.com/p/new-ways-to-corrupt-llms","https://consensus.app/papers/details/7cf9ed9aa21d5b85aaba765690c15d75/?utm_source=claude_desktop"],"category":"semantic","chatbot":"Other","strength":0.52,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"97a894c0-d95b-4b96-b884-534cbcabf880","slug":"pronoun-density-surge","title":"Pronoun Pinball","description":"So many it/they/this bouncing around that the reader has to guess; humans often name things more clearly in the same situation.","example_text":"It has been noted that they often rely on it to achieve their goals.","examples":["It has been noted that they often rely on it to achieve their goals."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://consensus.app/papers/details/37eb8cd604de548eaaf7e5b621caf478/?utm_source=claude_desktop","sources":["https://consensus.app/papers/details/37eb8cd604de548eaaf7e5b621caf478/?utm_source=claude_desktop"],"category":"linguistic","chatbot":"Other","strength":0.44,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"7c31ba6f-6453-40e6-a2ea-a80e08d133ef","slug":"slashout-rethink","title":"Slashout Rethink","description":"Shows revised intent inline (strike/replace) common in Slack-style writing; traces human editing, not a single polished pass. Stylometry work on human vs LLM text highlights editing traces as human-leaning (ACM Hypertext '24).","example_text":"Use the old ~~slow~~ fast endpoint.","examples":["Use the old ~~slow~~ fast endpoint.","Owner is ~~Jane~~ Jo after the handoff."],"tags":[{"key":"category:other","kind":"category","slug":"other","display_name":"Other"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://dl.acm.org/doi/10.1145/3703323.3703712","sources":["https://dl.acm.org/doi/10.1145/3703323.3703712","https://arxiv.org/html/2505.01800v1"],"category":"other","chatbot":"Other","strength":-0.42,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"cd9cdcc7-12dd-4c04-96c6-fe4233c2d153","slug":"viewpoint-balance-theater","title":"Viewpoint Balance Theater","description":"Everything gets the both-sides treatment: on the one hand, on the other, both perspectives have merit, served with polished corporate pleasantness instead of a real point of view.","example_text":"Both perspectives have merit and require careful consideration in the appropriate context. On one hand, this proposal has clear benefits. On the other hand, concerns remain.","examples":["Both perspectives have merit and require careful consideration in the appropriate context.","The service was exceptional and the experience was thoroughly enjoyable.","There are compelling arguments on both sides, and each perspective deserves equal consideration.","On one hand, this proposal has clear benefits. On the other hand, concerns about scope remain valid."],"tags":[{"key":"category:semantic","kind":"category","slug":"semantic","display_name":"Semantic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://consensus.app/papers/details/8b7f977a601c52cfb3d91d499dc22c7f/?utm_source=claude_desktop","sources":["https://consensus.app/papers/details/8b7f977a601c52cfb3d91d499dc22c7f/?utm_source=claude_desktop","https://consensus.app/papers/details/81dc9c6ef47c5d9f8a0d0acf783ae5e9/?utm_source=claude_desktop","https://consensus.app/papers/details/81b811255193527481c518aa333835f2/?utm_source=claude_desktop"],"category":"semantic","chatbot":"Other","strength":0.56,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"},{"id":"5c41b0f8-dff4-4fb8-b97f-d94d7bf44fbd","slug":"equivocal-qualifier-cluster","title":"However Habit","description":"Careful humans poke holes in their own sentences right away (although, but, sample size); flat confident AI paragraphs often skip that push and pull.","example_text":"The results suggest improvement, but the sample size limits generalization. Although the model performs well in distribution, it fails on edge cases.","examples":["The results suggest improvement, but the sample size limits generalization. Although the model performs well in distribution, it fails on edge cases."],"tags":[{"key":"category:linguistic","kind":"category","slug":"linguistic","display_name":"Linguistic"},{"key":"chatbot:other","kind":"chatbot","slug":"other","display_name":"Other"}],"user_name":null,"source":"https://consensus.app/papers/details/0d82bebd804d513db021a4cc5cff4b85/?utm_source=claude_desktop","sources":["https://consensus.app/papers/details/0d82bebd804d513db021a4cc5cff4b85/?utm_source=claude_desktop"],"category":"linguistic","chatbot":"Other","strength":-0.55,"community_votes":0,"created_at":"2026-04-05T01:55:04.69499+00:00"}]}