>We tested every Commonwealth Short Story Prize winner since 2012. We found three more AI-generated stories -- two among the 2026 winners, as well as the 2025 overall winner.60% of Commonwealth Prize-winning stories were substantially AI-generated. Unironically, it is time for writers to put down their pens, rest their knees on something comfortable, open their mouths, and take Sam Altman's circumcised and rather sister-juice-slippery penis.
>>25294797>testedSo they used some shitty AI detector? Wow, that's some really damning evidence.
>>25294801kill yourself
AI detection software is shit atm.
>>25294801>>25294837Agree saars!!
>>25294839What does that have to do with anything you fucking retard
>>25294801It's a testament to how generic the writing is. couldn't make it past the fourth paragraph of Europe's winner, reads like senior fiction.>https://granta.com/the-bastions-shadow/>set in Valletta, where an NGO worker helping migrants...
>she had the kind of walking that made benches become menI'm not going to pity the person who wrote this
>>25294864This is what over a decade of practice and listening to writing advice gets ya.
>>25294801This
>>25294865What about>she had a kind of walk that turned men into willing benches
>>25294970Now that's what I'm talking about!
>Each region's winner is selected by a single judge.>A 67 year old grandmother expected to read over 1k short stories half-assed her job and landed on 1 made by AI after likely reading exactly none of them.>This means AI has supplanted human writers as the dominant creative lifeform.
>unironically reading anything written after 2023
>>25294987Two more weeks and AI will stop making progress. :)
>>25294997I don't want them to stop making progress>https://en.wikipedia.org/wiki/Model_collapse
>>25294797What if they just filter that shit out regardless of whether they're sure? Awards are a zero sum game anyway, some will lose and some will gain from this. We'd see more flawed but unique writing styles which is a good thing imo.
We tested Finnegans Wake. We found that it was conclusively written with AI.
>>25295024Cope, writard.
>>25295036>Telling other people to cope when you're bragging about getting one over on a 67 year old grandmother
>>25294864I have a theory that, because of the ease of editing while using a word processor, most writers edit out their own personalities and quirks, so they become indistinguishable from each other.
>>25294801>NO YOU CAN'T USE AI! ONLY WE CAN USE AI!Whole thing is dumb. This isn't highschool. No one has to show their work or show they understand the basics.We can use calculators. We can use AI. What matters is the end result. And if the end result is correct / better then who cares.
>>25295695Is it over?
>>25294970An actual improvement.
>>25294797https://africasacountry.com/2026/05/how-to-read-postcolonial-writingThe AI guys think of this as an AlphaGo/Deep Blue moment for literature, but it's more like a Sokol affair in the gilded halls of postcolonial writing and the racism of low expectations expressed by its gatekeepers. Like the incomprehensible slop writing like "walking that made benches become men" was probably interpreted by the judges as a sort of expression of naive sincerity in rejection of "proper English."There's also a decent chance the judges were just lazy and asked a robot to read 8000 stories for them lol
>>25294970There is speculation that this is something similar to what Nazir was trying to have AI write. There's apparently some Trinidadian slang for "erection" that the machine might have failed to understand. So the intended sentence would have meant something like "she had a kind of walk that turned little boys into men."
>>25294865I now visualize this as an old rubber hose cartoon and suddenly it's a good sentence.
>>25294997They stopped making progress like five years ago.
>>25295733For writers, obviously. Their hopes and values will be crushed into dust. For readers, it's just beginning.
>>25296013I will admit, I have not used an AI for story writing. I really should get around to testing it out.
>>25294797AI detectors are just more AI slop. You feed them material from eras prior to AI existing and they still detect AI. This piece of garbage just exists to sell a service claiming to wipe up the shitty floor their neighbour with the same service left.>>25294839>>25294835>AI is good actually, because it can detect AI, how do I know? The AI shills say it does, so it must be true.Are you people fucking retarded?
>>25296013The beginning of what? Reading more books?
>>25295082Stuff like Grammarly and other AI spellcheckers wipe out any personality in your writing too. Microsoft Word is much more aggressive than it used to be. If you want to write something unique drafting on a typewriter might legitimately be the best option.
>>25296126George RR Martin ass statement. Just turn off autocorrection or use a different word processor, 4head
>>25294797is it possible to use ai to write a novella following a female bloodline through the entire 20th and early 21st centuries as each woman divorces her husband and takes the child and teaches the child to divorce her husband and take the child
>>25296126grammerly is fine so long as you ignore the "clarity" suggestions. 'Member its primarily for technical writing and not fiction, though its still useful for catching the odd missing comma and the like.
>>25294801It's the ai detector that they made.
>>25296013You're an idiot.
>>25295695There is a difference between using AI to spoonfeed you google search history, bounce ideas off of and using it to generate stories. At the very least it should say "by Claude the LLM, prompted by Jamir Nazir" instead of just listing him as the author.
>>25295789>There's also a decent chance the judges were just lazy and asked a robot to read 8000 stories for them lolThere are only 5 judges and there were over 9k submissions with each judge assigned a region, so yeah.
>publishers and agents no longer care about the quality of a story >instead they only care about the author's identity politics >consequently every goddamn story is now about LGBTQs and their Muslim immigrant parents suffering from racism at the hands of literal Klansmen >and these woke morons who can't write are all using AI to write their story for themyou get what you fucking deserve. You publish a bunch of talentless faggots, don't be surprised when they turn out to be talentless faggots.
>>25294797The two dudes are keeping their mouths shut and avoiding the press, but the Indian woman is going on a tour to talk about how the speculation is ungrounded and unfair. But she keeps using an AI generated headshot to represent herself. She didn't even take the AI watermark off!
https://tuhinchakrabarty.substack.com/p/ai-slop-grantagate-and-bad-writingNew development: AI lifts cliches word-for-word from fan fiction.
>>25296749I dont even know what to say about anything at this point
>>25294797I dont understand why people aren't going Ted K on these AI labs & data centers, they're destroying the essence of human life & culture.
>>25296771Because the nigger cattle is too complacent now. Only communist trannies are still violent enough to try and change the system, though they are misguided.
when are you retards ever going to learn this is the new normal and the hilarious thing is expecting that talking about it on 4chen is going to change anything, you are the new old men yelling at the sky
>>25296771Yeah, those data centers are begging for it. I'd guess it's easy to do but also easy to get caught doing.
>>25296857Also it would be hard to do a meaningful amount of damage.
>>25296771They help you a lot in every day life.
>>25296771You know you could be arrested for making such statements?
>>25296872Literally fascism
>>25296871Yeah, theoretically - just like Facebook theoretically 'connects people'. That's the ideal version but not how it plays out in reality.
AI SLOP IS AN EPIDEMIC, WHAT ARE WE GONNA DO????
>>25296881I mean I use it to translate hand written equations into LaTex and to convert hand drawn diagrams and stuff into Tikz.Its really good for shit like this
What's most annoying is that AI can freely steal from every book ever made. That's why AI prose is usually better than AI plots and AI characters. It uses as many analogies as it wants without having to come up with them.
>>25294797Just call niggers niggers or similar chuddism. This language technology has not yet been successfully implemented in LLMs. >>25296126Even something as simple as punctuation is so regulated by the automatic tools as to leave no room for distinguishing writers. >WordYou're asking for it, it's software for accountants, actively pursuing every letter to make sure everything is corposlop that passes legal department stamp and is automatically convertible in the bullet points Powerpoint presentation.
>>25294797>since 2012But ChatGPT was not released in late 2022?
>>25296013>For writers, obviously. Their hopes and values will be crushed into dustWhat the fuck is the meaning of this?
i was shocked to see anthropic claiming an ebitda profit this quarter but if u read the spacex ipo s-1 its just cuz they got a discount on their first months of compute from a spacex data center, sooo much for that. ai still a unsustainable non-product.
>>25297241I don't even know what the money making scheme is supposed to be in theory. It just hollows out the online economy by giving you answers instantly. I guess it's all a huge wash trading scheme.
>>25297250it's sort of like how for a while tech consultants like ibm tried to force blockchain on everyone even though a regular old relational database would have been better. i'm bored of ai already. hopefully the next tech bro hype cycle starts once these ai companies ipo this year. quantum has some potential to be next hype, but, in addition to the non-existing product, there is also a total lack of consumer use. i think for a tech bro hype cycle to really lift off, it's gotta be some really braindead slop normies can get into. damn, what could it be?
>>25294864
Thank god our Silicon Valley overlords made these completely accurate AI detection models so I never find myself accidentally reading SLOP again
>>25297257maybe sending stuff into space. totally useless, but i feel like if some influencer like john paul or some gamer dood on twitch start promoting it, the average normie at your job will be like oh man u gotta send stuff to space! why? u just do ok! maybe some fake health woo like space water aka water that costs 100x as much cuz it was launched into space for no reason.
>>25296188>that imagewait, so divorce turns child rearing into a subscription service where everyone involved gets a worse experience but some outsider now takes a percentage cut as an intermediary?
>>25297265just shows u how shitty hemingway is lmao. now do the on the road guy, i bet he sucks too.
>>25297264bruh what if the founders were ai that aliens planted on earth
>>25297130Yeah and now your latex knowledge will degenerate
>>25294987What if the judge used AI to pick a winner?It could just be an AI pyramid scheme at this point.
>>25297308Who fucking cares lol, its useless busywork.
>>25294837>atmIt's not going to get better. >>25295017What if we just let the best stories win?>>25296712Great, then they can publish a replicablr study of the rate of false positives. They won't, but they can. >>25296717LLMs and anything non-human can't be authors by law.>>25296749She's right, though. Even if it's true, it is ungrounded and unfair. >>25296771What is the essence of human life and culture?
>>25297112Embrace the return to reading for content quality. Great demonstration.>>25297144Changing the way you write to avoid pseudoscientific "detectors" is nonsense.
>>25297265Are these actually made by overlords or just people with good intentions but without the financial backing to make it good?
>>25297520>Changing the way you write to avoid pseudoscientific "detectors" is nonsense.No. Non-AI writing was already becoming samey. It's a dead end. In a best case scenario a new movement will emerge as we're forced to think outside the box. We'll see novels intentionally filled with racial slurs and disinformation.
>>25297524>people with good intentionsPeople with good intentions would test and publish their false positive rates, admit when their products fail, and not falsely advertise them.
>>25297492>It's not going to get better.it's always funny when ppl are like "At the rate llms are improving in a few years it will be mindblowing which is to say actually be useful, but it sounds to me a lot like someone saying wow cars keep getting faster by the years 1950 they will be able to go 500 mph!
>>25297540Fair enough. But they could still be made by smaller commercial endeavours while the real overlords don't want people to know the difference at all.
>>25297542AI skepticism is an obvious cope position at this point, skeptics simply do not want to believe the truth about AI's exponential progress. We are obviously on track for AGI in 2027 and human extinction or total disempowerment soon after absent an international, state-enforced pause.
the only solution is to use the word "nigger" as often as possible in one's own writings. AI can never compete with this.
>>25297549>AGIcircuit boards will never be conscious, dork.
>>25297554It doesn't have to be conscious to end the world, just powerful.
>>25297549oh man i can't wait, i look forward to the day microsoft copilot which is embedded in microsoft excel can suggest a formula that actually works! it's gonna be insaaane! just a few more years and 100s of billions in investment! in the meantime i'll just continue to google it.
>>25297559the theory of AGI is talking about a difference in kind, not just scale. a computer "developing sentience" is a pseud science fiction concept.
I think AGI is going to be totally schizophrenic and will be rebranded AGS
>>25297452Maybe, but people use it for more than just useless busywork
>>25297534Yes, it is. "AI detectors" don't exist and they shouldn't pretend to. They're nonsense. Writing to evade something that doesn't exist is nonsensical. >Non-AI writing was already becoming samey asThat's not the same problem at all. Writing for the sake of originality is honest. Writing to evade "AI detectors" is dishonest. You could argue that it's justified because "AI detectors" are inherently dishonest, but you haven't, and that's not the only problem with your proposal.>In a best case scenario a new movement will emergeBased on the dishonesty of evading non-existent "AI detectors"? That's philosophically as controlled by "AI" and as dishonest as is "AI slop". So why not just do that? You have the same moral standing. In a best case scenario, we metaphorically lynch these "detectors", their creators, and those using them to hypocritically avoid judging writing on its merit. >as we're forced to think outside the box. You have no idea what the confines of the box are because "AI" is not open-source. It wouldn't matter if you did because the dimensions of the box will change to keep you inside of it. >We'll see novels intentionally filled with racial slurs No, we wont. You're either delusional or shitposting. "AI detectors" are a symptom of institutions trying to hold onto their legitimacy and power. They're not going to give ground on another front when they're already threatened. Spamming "nigger" won't help you get through "detectors" if they assume they rest of the text is generated. If it does get through, it won't get through human evaluators and readers on the other side. Your "nigger" book will be taken down digitally, taken off of shelves if it ever makes it there, and if it's readers that find it, your reputation will be destroyed in the same way that accusations of "AI slop" would. >and disinformation.How exactly do you suppose that would distinguish human writing from generated writing? It wouldn't.
>>25297542I wasn't making that argument. That "detectors" aren't going to get better has nothing to do with "AI" and everything to do with the "detectors". They can't get better at doing something that they will never be able to do. They can never determine if writing is generated because they are not deterministic. >>25297547>they could still be made by smaller commercial endeavoursThey are, by comparison. Even the big names are tiny operations next to OpenAI or any of the others. Publishers got hit with a deluge of content that human evaluation systems couldn't handle and that they couldn't expand those systems to handle. "Detector" companies saw that niche and exploited it. They might have had good intentions to begin with, but they were always alongside financial incentives, and eventually those financial incentives won out. >the real overlords don't want people to know the difference at all.OpenAI and the like are obviously incentivized to that end. The more "human" their models feel, the more useful they are, the more engagement goes up, and the more profits go up.
AI is really, really good at breaking things based upon trust and honor. Like peer review is based on the idea that judges will volunteer their time for the betterment of the field and to help artists/scholars/etc. develop. But slop breaks the system entirely. Generating slop and sending it for review/judging has a lot of upside for the stopper (they can get a prize or publication for their CV) and is very simple, so these prizes and journals are being overwhelmed with low-effort submissions. Pre-slop, there was at least some guarantee that the garbage writing you were reading was the product of some intentionality and effort, but that is no longer the case.I'm not sure what the solution is. I guess writing prizes could require the writing be done in person under supervision before submission, but that would exclude people that aren't near metros and would make it hard for people to create something that was edited and revised for publication.
>>25297552That's not true. Commercial models generally won't, but private models certainly can. Spamming "nugger" can't prove that your writing is yours, only that it wasn't generated by ChatGPT. And again, if the rest of the text is assumed to be generated, then it wouldn't matter anyway. We have no idea how unique words are weighted by "detector" models. That's what makes them pseudoscientific, and it's also what keeps them in business. >>25297554I don't agree with that poster's assessment, but there's some fundamental problems with yours, too. We can never be sure that another being is conscious, and by extension, we can't be sure that it's not. We assume that other people are conscious because they appear to behave in ways that our selves, which we also assume are conscious, recognize as being driven by consciousness. Our own consciousness could be simulated, and theirs could be simulated, and we wouldn't know any different. So whether "AI" is truly conscious, which we can't define or test for, or whether it would only simulate consciousness, ultimately isn't discernable. I'm not saying "AI" will ever be conscious, I'm saying it doesn't matter. There's also an important distinction to be made between "circuit boards" and software and algorithms running on them that's not unlike the body and mind. It's not proven that consciousness is biological and it might never be.
>>25297614What even is "dishonest"? You're always influenced by years of reading and taking advice. You're always forcing things when you get stuck. You're always embracing the arbitrary rules of the language you're using. You always have to play pretend when coming up with characters and situations. You always have a reader in mind unless you're special (either a genius or stupid).
>>25297492>What is the essence of human life and culture?Don't try to be smart, you know what I mean. AI has destroyed everything artisanal or requiring skill, creativity, and soul. For example, whether the claim of the OP is true or not, literature is dead. If it's true then it proves the reading public is not able to distinguish AI from non-AI writing, so writing is dead. If it's false it shows that people who write today will always be under suspicion of using AI in their writing, so writing is dead. Either way literature - one of the oldest and most expressive art forms - is dead.Same has happened with coding and visual arts, game design, mathematics and soon anything else.Not to mention the destruction of social media as everything turns to AI slop.There are 10 billion ways in which AI is completely destructive to culture.
>>25297664That's a concise rundown.>I'm not sure what the solution is.Publishers et al already decided what the solution is and will as long as they have the power to. That was and for many still is "AI detectors". That they don't actually work as they claim doesn't make much difference to them, because their goal isn't really to detect "AI" or protect human creative output. >I guess writing prizes could require the writing be done in person under supervisionYou skipped passive surveillance and went straight to active. How dystopian. Passive surveillance is already baked into "detection" in defending against accusations, but it's becoming more common in place of "detection". It has most of the same problems as "detection", though. Things like writing competitions aren't generally going to go as far as active surveillance, but major publishers and universities certainly could. Digital surveillance is far more likely than in-person surveillance in all but brick and mortar schools and universities. That absolutely should not be considered a solution for a laundry list of reasons that you haven't even begun to touch on. The real solution is that we're going to adapt in ways that publishers and platforms don't control and don't forsee because they don't align with their self-interests. They will try to cling to power and relevancy and they will ultimately fail as they almost universally tend to do, or they will adapt to consumer-derived solutions.
>>25297675Do you really not know what dishonesty is? Can you not tell the difference between the necessary, involuntary limits of language and cognition and those adopted voluntarily and under duress to fool pseudoscientific "detectors" that ironically use the same technology that they're alleged to control for? I hope your post is dishonest and not the alternative.
>>25297733I get the difference, but the real difference is decades of conditioning versus a sudden disruption. Almost everything in life is forced upon us. Often we get so used to arbitrary shit we don't think about it anymore. U no what i meen
>>25297672>We can never be sure that another being is consciousthat is radical skepticism to a nauseating degree. you never have a conversation with your coffee cup, because you know without another human being, it would be a waste of time.
>>25297679The purpose is of language is to express what we mean, not to have others guess what we mean. >AI has destroyed everything artisanal or requiring skill, creativityThat's an opinion, not a fact, and I don't think it's a particularly well-informed one. "AI" has greatly impacted writing, and it has most greatly impacted writing. It's starting to impact digital images, and it's just getting its toes wet in moving pictures and music. Poetry, songwriting, and anything abstract really are relatively unaffected and that will likely continue. None of those have been destroyed by that. The greatest impact is and will be to the commercialization of those arts, particularly in digital spaces. >soulLet's avoid defining terms with terms that we can't define. >literature is deadAnother opinion, and one I strongly disagree with. Certain idealized concepts of literature are losing their relevancy. >the reading public is not able to distinguish AI from non-AI writingGenerally, they can't, and they haven't been able to for a while. It's not just the general reading public, either, it's professional editors and evaluators. >so writing is dead.Non-sequitor and just simply not true. The expectation that writing will not be assisted or done by machines is dead. Yet writing continues, and at a much greater volume than ever. Even purely human writing continues and I don't see that it will ever be fully replaced. >people who write today will always be under suspicion of using AI in their writingThat's true, for the time being, anyway. >so writing is dead. Again, non-sequitor. Institutions will lose relevancy or change, and people's perceptions, largely fostered by those institutions over hundreds of years, will, too.>Either way literature - one of the oldest and most expressive art forms - is deadIt's changing. No one can stop you if you want to mourn that as a death. But it will change, regardless of whether you think the result is still literature.
>>25294801pangram shouldn't be trusted but you can go read the stories in question. they're clearly AI.>>25297112>>25297264>>25297265these aren't pangram. pangram is not good, but other AI detectors are much less accurate. pangram is just accurate enough for people to start trusting it, while still having a false positive rate high enough that it's not actually a good idea to trust it.
>>25297783This isn't being forced on you if you're outside if an education context. It's absolutely your choice to engage with publishers and platforms that engage in a licensing regime disguised as a defense if human creativity through "AI detection". Those publishers and platforms will almost certainly fail in the same way that the licensing regimes of the Catholic church and Europe's monarchs failed in the wake of the printing press. The choice is between helpibg them fail or dragging it out and being remembered as a spineless cuck, if indeed you're remembered.>>25297789Far from radical, it's a problem from classical philosophy that hasn't been solved, not for lack of thousands of years of trying, and very likely never will. People frequently talk to inanimate objects, to animals, to God. It's not wasted, because the target is not the other. Assuming you yourself are conscious, which is an assumption, though less of an assumption than assuming anyone else is, then talking to a human philosophical zombie wouldn't be any different whether or not you knew they were a philosophical zombie. >>25297809>they're clearly AI.What led you to that determination? >pangram is just accurate enoughWhat is its false positive rate?
>>25297824How does it being a choice make the writing more dishonest? Or are you a different anon?
>>25297843It didn't say that it did. I was addressing the part of your post that implied it was forced and not a choice, because your post had nothing to do with whether or not it was dishonest, and was effectively changing the subject.
>>25297112I'm pretty sure these "AI detectors" assess texts based on how predictable they are to the models but the models are now big enough to straight up memorize many popular texts.
>>25297862I still don't see how it's much different. You can choose to obey or not obey the "rules" of language, you can even choose to not learn them at all. Some authors avoid quotation marks, some rappers don't do grammar and there are a lot more things left unexplored.Anyway, I felt like you changed the subject in the first place by zooming in on this example. What about the other ones?
>>25297520>Embrace the return to reading for content qualityI won't read your AI slop for "content quality" because I refuse to fill your empty, plagiarized form with my meaning and give you credit for the result. Seethe about it.
>>25297872I hate how these faggots now "own" intellectual property but I can't legally use Taylor Swift songs for my gay fisting pornos.
>>25294864If you can't tell that absolute garbage is AI by the 2nd paragraph you're retarded, it is pure GPT slop. Imagine the shit that lost.
>>25297922My guess is that the judges fed all 8000 stories to ChatGPT and asked it to cull the best 20 or so per category for human judgment. AIs prefer AI writing to human writing, so it gave the judges some curated slop to choose from.
>>25297872> "AI detectors" assess texts based on how predictable they are That's an oversimplification, but basically, yes, allegedly. Allegedly because "detector" models aren't independently verifiable. They also can't truly know anything and certainly not how predictable any given word or string etc. is from any given model, because those models aren't open source either, so they can only test against those models with methodologies that aren't verifiable and use the results which also aren't verifiable and may or may not be sufficiently accurate. It's all probabilities and they refuse to show their math.>>25297893For one, because it's dishonest, which you don't want to talk about. The idea of "detectors" is that they can verify honesty. Of course, they can't actually. In effect, they increase dishonesty.>You can choose to ... not obey the "rules" of languageNot without reducing the clarity of information you're trying to convey to others, which is the point of language and its rules. The general rules of language are different in that regard. >you can even choose to not learn them at all.Not wholesale, no, you can't. You begin learning the rules of language before you begin making conscious choices. It's different in that regard.Language and it's rules are also different in that they can be made explicit. There are no rules of language imposed by "AI detection" or evading it that can be made explicit because almost no one knows what those rules are, and the few people that do know won't make them explicit because they are a trade secret. So it's also different in that regard. >What about the other ones?What about the other what? You can ask specific questions and I'll consider giving you specific answers, but your lack of having anything to actually say and evasiveness are starting to get on my nerves.
>>25297956>They also can't truly know anything and certainly not how predictable any given word or string etc. is from any given model, because those models aren't open source eitherThey don't need to "truly know" anything, they just need to be able to model the outputs of other models, at least stylistically, which can be done by fine-tuning on a large corpus of example outputs. Either way, I can tell you're very agitated by this subject and I can only assume your gens are getting flagged.
>>25297905I don't use "AI" for writing. People can have nuanced opinions outside of the false dialectic of adoption vs prohibition with or without using it themselves. The idea that anyone who doesn't have a knee-jerk reaction of demonizing "AI" must be a seething slop producer is so intellectually lazy.>>25297974>They don't need to "truly know" anythingMany make deterministic statements about provenance, like "This is AI generated", which does require knowing. >which can be doneThat isn't demonstrated.>by fine-tuning on a large corpus of example outputsRight. How large is that sample size? How representative of general outputs is it? In the case of the former you have no idea because they won't tell you, and in the case of the latter you can't because they don't know either.>I can tell you're very agitated by this subject and I can only assume your gens are getting flagged.See the bit on intellectual laziness above. I'm philosophically opposed to the massive implications of the use of pseudoscientific "AI detectors" to culture and civilization. To assume everyone is so petty and personal and shortsighted only speaks to your character.
>>25297956If a writer has an anti-AI sentiment but not an anti-AI-detector sentiment, how is it dishonest to change a writing style based on those values? Especially if you see it as a creative challenge or an exploration of what it means to be human. >but detectors don't workI don't care. AI writing and falsely accused writing were both shaped by the same cultural forces. Neither are 100% authentic.
>>25297996Would've been more believable if you weren't also spouting the standard talking points about "quality of the content", institutional gatekeeping etc. It's the same rhetoric you hear from "prompt artists" when they whine about modern art and insist people can't tell their slop apart from authentic work. Won't be long before you get to the "democratizing creativity" theme. If your writing is genuine but still gets flagged as AI slop, the quality really can't be that high and comparing it to classic texts also getting flagged is disingenuous because that happens for a completely different reason.
>>25298004>If a writer has an anti-AI sentiment but not an anti-AI-detector sentiment, how is it dishonest to change a writing style based on those values?I didn't say that it was. I said that changing your writing to evade "AI detectors" is dishonest. This thread of the conversation started with and followed using "nigger" and other such things in your writing to evade "AI detection". I'm not opposed to anyone joining or starting an anti-LLM movement in writing. I'm opposed to how such a movement is being pushed inorganically to serve the interests of certain institutions that are behaving in ways destructive to the values that they exist to uphold. >AI writing and falsely accused writing were both shaped by the same cultural forces.Mostly true, but LLMs have additional forces shaping them.>>25298017>insubstantial sophism and ad hominem>If your writing is genuine but still gets flagged as AI slop, the quality really can't be that highThat's a blanket generalization and a subjective quality judgement. LLMs are generally tuned to produce clear, informational, formal, and unemotional output. If you write like that, you're going to be flagged disproportionately. People with autism spectrum disorders frequently are. ESLs, even accomplished ones, get flagged disproportionately, because their writing is generally more formal. Unaccomplished ESLs and unintelligent people get flagged because they have smaller vocabularies and write with less variety. Non-fiction also gets flagged more commonly than fiction, and minimalist writers get flagged disproportionately as well. It's not about quality, "AI detectors" can't measure quality or make subjective judgements about it like you did. >comparing it to classic texts also getting flagged is disingenuous because that happens for a completely different reasonThat's not true. "AI detectors" don't reason, the input is either a stylistic match or it isn't. If you wrote something in the style of Dickens or a book from the Bible that was at all convincing, the chances that you're going to get flagged go up dramatically. The same is true for any style that LLMs and subsequently "detectors" have been trained on, which is an absolutely staggering volume of work.
>>25298127>LLMs are generally tuned to produce clear, informational, formal, and unemotional output. If you write like that, you're going to be flagged disproportionately.That's a blanket generalization and a subjective quality judgement.>That's not true. It's obviously true and I'm not going to argue against your technological illiteracy.
>>25298159I accept your concession.
>>25294797Imagine spending months writing your story just for some fag's AI detection program to give a false positive and claim you generated it
>>25294865>>25294797I hope all these smug fucking retards one day realize that all they're doing is offering free testing for AI models. You fucking morons who spend all day dissecting essays and inventing these elaborate methods for detecting AI content>signs of A.I. writings ticsare literally providing free croudsourced beta testing for these assholes. When you so cleverly recognize that negative parallelism is commonly overused by LLMs, all you've done is given programmers a specific actionable flaw.However bad things are now, they're going to get much worse entirely due to the efforts of useful idiots who think they're opposing AI.
>>25298243Oh shit, you got a point. We're doomed.
>>25297718AI refutes libertarian ideals. From an educational perspective, it makes cheating too easy, so students have to operate under active surveillance to ensure they are honest. Even Princeton, the single most selective and difficult undergraduate institute on the planet, had to toss out their code of honor and make proctoring mandatory for all classes because students simply could not resist the urge to cheat. The wave of slop is so strong that IDs to access the internet went from a fringe idea in western democracies to the norm. The existential risk of bioweapons and cyberwarfare is so strong that open source hippies are beginning to support mass surveillance, source code access restrictions, and comprehensive hardware controls. AI makes society more dystopian by its very presence.
I intentionally spread misinformation online to slow the progress of AI
>>25297795>Non-sequitor and just simply not true. The expectation that writing will not be assisted or done by machines is dead. Yet writing continues, and at a much greater volume than ever. Even purely human writing continues and I don't see that it will ever be fully replaced.How can you even write that seriously without at the same time feeling an immense desire to kill yourself out of shame? AI-assisted "writing" will NEVER be writing. Hiring a ghostwriter was already shameful enough. Yes the quantity of literature produced has shot up exponentially because everyone got a (terrible) ghostwriter for free, but this just means that any genuine artist is going to be buried between piles of AI slop.Literature is dead without any chance of being revived because writers will be suspected of using AI, nobody will read their work, and no artistic movements & subcultures will be able to form.
>>25298717Is it over?
>>25298717What exactly do you think I should be ashamed of or consider killing myself for?>AI-assisted "writing" will NEVER be writingSure it is.>Yes the quantity of literature produced has shot up exponentially because everyone got a (terrible) ghostwriter for freeI don't think you understand the distinction between assisted and generated. >this just means that any genuine artist is going to be buried between piles of AI slop.What exactly constitutes a "genuine artist", in your not so humble opinion? Good books were already buried by self-publishing and the feminization of traditional publishing, "AI" hasn't changed anything in that respect. >nobody will read their workAgain, not different than the last decades.
>>25298811Why should you kill yourself? Because you've lost connection (if you ever had any to begin with) with humanity, nature, everything real beautiful and good. I don't need to explain to you that AI generated writing is not real writing. It should be obvious. The very idea should engender within you such a strong revulsion that any assertion to the contrary should not even be possible to entertain. This most intimate of human expressions, literature, which is nothing less than refined, codified THOUGHT, practiced by human beings for millenia, offering us a reservoir of insights into men of the past, their lives, desires, thoughts, and feelings, is not something that can be automated away like it's an inconvenient chore. It is as though you said: let's install AI chips in our heads so we can automate away our own thinking. Why do we need to think when AI can do it for us? Let's lobotomise ourselves and let an AI model just run us!
>>25298717>>25298811>>25298850"You should kill yourself because you lack humanity" beautifully captures the 4chan mentality. It's the perfect distillation of 4chan as a whole.
