38 pointsby theanonymousone9 hours ago21 comments
  • throwa3562624 hours ago
    I am proposing a new law: Any content where more then 2% has been created by AI should be clearly marked as such.

    Would improve both Amazon and YouTube significantly. I have basically stopped using them after seeing all LLM spam.

    • Eddy_Viscosity23 hours ago
      No way to enforce it. We already have laws about scams and counterfeit products which Amazon does not enforce, even in cases where it absolutely could.
      • zbentley3 hours ago
        There are valid objections to GP’s point, but this one is silly.

        Having anti-counterfeit laws is vastly preferable to not having them. Counterfeiters do get busted, though you’re right that Amazon and the authorities could indeed do more.

        I’d rather live in a world where at least some criminals are deterred from selling me a fake fire extinguisher that doesn’t work when my house is on fire.

      • throwa3562622 hours ago
        No, but once you start demanding your money back for a product being mislabeled things will improve
  • impostervt4 hours ago
    Amazon KDP recently put a limit of 2 books published per week because of this kind of thing.
    • lp9235 minutes ago
      Seems like it should be 2 or 3 per month.
    • thewhitetulip4 hours ago
      It takes me 1 year to write a book (fiction and non fiction alike)

      2 books per week is still absolutely insane

  • skeledrew3 hours ago
    > "Books are selling."

    That said it all, and literally the only thing that matters, because sales means demand, which means people find the works valuable.

    • beej71an hour ago
      Buyers think they'll find the work valuable. We've definitely seen publicized examples of buyer's remorse when an LLM-authored book is cracked open.

      If someone's just in it for the money, why wouldn't they slop it up?

      • skeledrew42 minutes ago
        If people are remorseful enough then there'd be enough feedback in reviews that others know to beware. And that would lead to author names associated with those works being outed and likely boycotted or something. This particular author has ~11 pen names, so assuming a roughly equal spread across her 370+ books, there should be a crop of bad reviews for books associated with a name that has maybe 25 - 40 other books associated with it. There should be a pattern for people to find and then say "Someone has been putting out slop under these names; avoid like the plague", which should then trigger a considerable gap in sales for that author.
  • 4thguy3 hours ago
    I'm unsure where the line is drawn for LLM ect in the arts is supposed to be, but the article does say that she "heads an operation that has trained several hundred people in her method". So, let's take the LLM out of the question, is it fair to say that she is the author?

    I'm not saying that authors should mine clay, hand-carve the letters on a clay tablet, and fire it up in a kiln that they built themselves; but authorship requires intent.

    • mzajc3 hours ago
      From the figure in the title, she spends under 2 days per book. That's from the start of her "career", so I assume the rate has been ramping up. This isn't enough time to read and review a book, let alone write one, so it would indeed be delusional to call her an author.

      She figured out a way to make a quick buck off gullible and uncaring people, and a market that has not been thoroughly saturated with slop yet. There's not much more to it.

  • netdevphoenix4 hours ago
    Not sure how her nationality is relevant here. Only the usage of LLMs in this manner matters
    • yorwba4 hours ago
      It's a local-interest story by a South African news outlet.
      • netdevphoenix2 hours ago
        Even more reason not to include her nationality then.
  • origini4 hours ago
    We have books pre LLM writing about 5G, memories of Water and other garbage.

    A book was never a quality criteria.

    The community, bestseller list (easy to manipulate) etc, were.

    Its the same with the critisism about Steam and Indiegames.

    • Eddy_Viscosity23 hours ago
      The problem is scale. Human-produced garbage content come in at rate X, but LLM produced garbage can come in orders of magnitude faster because you can spawn off agents to produce garbage content multiplying the output. The only thing limiting this is the economics and since the cost to produce this garbage can be so low, the amount that can be produced while still be economically feasible is massive.

      It's the difference between a regular ocean wave and a tsunami. It's the difference between a person bumping into you and a being hitting by a speeding semi-truck. Scale is the problem with LLM garbage content.

    • lopis4 hours ago
      What's the end game? Once we have a billion books published online each day, how can a legit writer with something real to say break through? Same with indie games. I'm in a game jam community and we're struggling with this. The sudden influx of entries (50% increase this year), mostly of low effort copied games, puts an unfair burden on everyone who needs to judge games - a fundamental part of how game jams work.
      • origini38 minutes ago
        you increase the requirements.

        Steam wants 100$ from you, tax documents etc.

        After that, you make the initial review very cheap for you: YOu need to create a promovideo from ingame.

        If this is still too much, you make a pre sampling phase. Everyone already developed a game gets more karma points and is easier found, everyone new has to earn their standing.

      • yorwba3 hours ago
        If you want to judge more entries without increasing the work for the judges, you need to scale the number of judges with the number of entrants, i.e. use peer review. Of course if most entries are low-effort, you'll get a lot of low-effort reviews. But if you have an inner circle of trusted judges, you can have them look at the winners according to peer review and, if they don't meet the bar, throw out reviewers who rated them highly. Repeat until you've found a subset of reviewers who are taking the task seriously.
      • skeledrew2 hours ago
        > a legit writer with something real to say

        What does that mean? What defines a "legit writer" and who determines when they're saying "something real"?

      • profunctor4 hours ago
        Critics, recommendations from existing known authors, publishers being clear about what is and is not ai generated.
        • jasomill3 hours ago
          This. Ironically, an increase in slop production could end up making it harder for new non-AI authors to get recognized, due to default assumptions that unknown quantities are slop.
      • sandworm1014 hours ago
        >> mostly of low effort copied games

        All AI games are heavily bloated. Try cloning Factorio or Prison Architect. AI can make something that looks similar to such games, but the back end falls apart.

        I have friends who keep showing me their new "games". I just see some version of Doom, but one that needs multiple GB of memory to accomplish what was commonplace in the 90s with a millionth as much headroom.

    • soon_to_be_sb4 hours ago
      >writing about 5G, memories of Water and other garbage.

      Have you actually read some of these garbage? If not, how do you decide it was garbage?

      • origini4 hours ago
        Yes I did.

        I did not only read 5G book about how bad it is, i also read the 5G spec book.

        I read books about pendulum (different types of them too, like one with stones you can drop), tunnels and pendulum, energizing water, homeopathy (how you make it, the theory behind it).

        Why?

        • soon_to_be_sb2 minutes ago
          What did the 5G book talked about and why was it bad?
    • vintermann4 hours ago
      Yes, I think people are a bit quick to forget how much pre-LLM slop there was. In all sorts of media. Asset flips on steam, Chuck Tingle-level "romance", algorithmic spam on streaming services (think baby sleep rain relaxing lo-fi beats) etc.
      • morkalorkan hour ago
        Hey, I'll have you know that Chuck Tingle is self-aware satire-slop. It's on a different level than asset-flipping scam-slop.
    • Razengan4 hours ago
      Yeah, people suddenly got selective amnesia and act like "slop" never existed before 3 years ago.

      Hell, look at the video games from the 1980s. Download the scans of any early gaming magazine or a MAME ROM archive. At least 50% of the games were derivative slop. Crappy movies and books were around for a 100 years too.

      • cwillu4 hours ago
        No, people are rationally responding to a massive change in the economics of producing the slop.
      • soon_to_be_sb4 hours ago
        >act like "slop" never existed before 3 years ago.

        Volume matters lol..

        • jasomill3 hours ago
          True, but at some point it doesn't. Ignoring the 99.99% of titles on Steam I don't recognize from reviews, known developers, or word of mouth isn't meaningfully harder than ignoring 99%.

          This is worse for new entrants than either established developers or consumers, who will have to work harder to get their work recognized amidst the rising tide of slop.

          Perhaps some conventions will develop similar to what I've seen on a smaller scale for fan artists releasing work online, where developers will provide some evidence of their development work in process to convince (human or AI) slop filters to let them pass.

  • beej71an hour ago
    I can assure you that no writing of mine is LLM-generated.

    But I wish I had some way to prove that. (Live writing? Git logs—but a lot of my chapters see a big first commit? Some kind of PGP web of trust thing?)

    For one thing, the goal of writing for me is not to produce a book. The goal is to learn and help others learn. The goal is to make the most concise, best quality material that has ever been written on that topic for that audience. It's a lofty goal that I can't claim to achieve, but that's the target. LLMs appear to be light years away from being able to do so and are easily bested.

    And, perhaps most importantly, writing is fun, relaxing, challenging, frustrating, rewarding, and human. Having an LLM write something for me is none of those things.

    I wish AI would do my fucking dishes, though.

  • tsaixingwei4 hours ago
    There is an established market for sudoku books. So if a publisher using AI to generate more sudoku books with lower costs and at a faster rate and people still buy those books because they enjoy solving sudoku puzzles, what difference does it make if a human painstakingly made those puzzles versus an AI generating them?
    • zbentley3 hours ago
      Because sudoku is a math puzzle and a book of words is a story.

      Like, the line between big artistic/creative efforts and other stuff is subjective and hard to trace, yes. I know the counter arguments here (“what if the book is an instruction manual?”, “what if someone’s labor of artisanal love is making sudoku puzzles?”), but sudoku vs story is like disagreements over “what counts as the Midwest”: you know very well that Antarctica isn’t in it, and that a sudoku book merits very different appreciation and artistic protectionism-from-AI than a story.

      It’s fine to categorize things without having a rigorous definition of the category.

    • mzajc3 hours ago
      She is not generating Sudoku puzzles, so you can drop the hypothetical. As for the implication that good sales figures mean people enjoy the books she generates, might it be possible that the reason her books sell are because she stuffs storefronts with so much slop (see the figure in the title) that she outcompetes everything else by volume alone? Regardless of quality, whether people like her books, and with no regard for authors who use higher brain functions at the cost of spending more than a day and a half writing a book?
      • swed4203 hours ago
        It's unfortunate this needs to be explained.
  • RugnirViking5 hours ago
    a debate I had with a friend recently was about how there aren't actually that many people contributing to the spam everywhere of AI crap. There are tens of thousands perhaps, probably not even enough to fill a small city, spread around the entire global population.

    A lot of people use AI, some even are unlucky enough to post online before learning its a faux pas. Some few are oblivious enough to go for some time before realizing.

    A select few decide that this is their opportunity to industrialize production of slop for profit. It seems that when platforms take steps to curb excesses like people posting thousands of products/posts/books/games/videos/whatever a day, the volume drops dramatically, because its these villains posting thousands of times a day that are responsible for most of the nonsense

    and as with this woman, whenever they see a modicum of success extracting money from the commons, they immediately pivot to selling courses, so other malignant souls can clog up the shared processes of humanity with their crap too.

    • za_creature4 hours ago
      This is the true killer app for LLMs. A lot of important people didn't like a media channel they couldn't control and many businesses did not like having fair reviews of their products and services.

      First it was "bot" accounts from low income countries posting and rating to skew the numbers, now you can do the same with actual bots. Drown the signal in a pool of noise until nobody believes anything any more and we return to the status quo from before personal computers.

      Except with a surveillance apparatus that would've made the soviets jealous.

      • brador4 hours ago
        Trustable curation is the answer.
        • brendoelfrendo3 hours ago
          I've already seen it impacting community curators on Steam; their inboxes are overflowing with vibe-coded slop. Curation is important, but it doesn't really scale; it just shifts the need to be a "human slop filter" from you to someone else.
          • za_creaturean hour ago
            It also creates excellent incentives for corruption.

            TV is a perfect example of this: you only have so much screen time to allocate to the news, so which subjects do you "curate"?

            I bet LLMs would love to curate the internet for us from all of that slop.

        • za_creature4 hours ago
          Don't worry, they'll slopify that too. TikTok is great at this.
        • soon_to_be_sb4 hours ago
          Goodhart would beg to differ...
    • mrweasel2 hours ago
      Perhaps not selling courses, but pretty close: https://coralhart.com/
      • RugnirViking2 hours ago
        from the article

        > "The South African now heads an operation that has trained several hundred people in her method"

  • brendoelfrendo3 hours ago
    I object to her being called an "author."
  • mcc1ane4 hours ago
    "I Don’t Want to Read What You Didn’t Write"
    • skeledrew2 hours ago
      ... but if you can't tell that they didn't write it...
    • foxtrottbravo4 hours ago
      Very substantiated...
  • benrutter5 hours ago
    The number of books makes this hilarious. Whatever your take on AI generation in writing, it's pretty difficult to read 370 books in 20 months closely. Let alone consider, edit and improve.

    There's just no way you can claim "oh no, I have a full creative process in place to produced something considered", whilst turning out an unreadable quantity of books.

    • 4thguy3 hours ago
      I did the math when I was trying to write a comment in a similar vein (yours is much better worded). Assuming that months are a generous 31 days, she'd have to read 0.6 books every day to read them all. That's on top of everything else.
  • glimshe4 hours ago
    > ...Claude, "really good story structure, terrible at sex drives"

    This is the most disturbing snippet about AI I've read in the past few weeks.

    • skeledrew2 hours ago
      Why? If one gives it a charitable thought I feel like it's obvious she means Claude doesn't do well with convincing sex scenes (I'm guessing as much of that content as possible was removed from the training data and otherwise suppressed). Grok is where the really steamy stuff is at; no filters.
  • flr034 hours ago
    > "Readers don't read rubbish books; they return them," she said.

    Do people actually return books they don't like? I don't but maybe it's a thing. Or is it her trying to convince herself the books are actually good, or at least not rubbish? Which is putting the bar low anyway.

    • sevenseacat4 hours ago
      Nah, stuff like this all goes on Kindle Unlimited where readers pay a sub and authors get paid by the page.
    • skeledrew3 hours ago
      If it's cheap enough people likely won't bother to return them, but if it's rubbish then there will inevitably be reviews that'll lead to people avoiding the author like the plague. That's the real signal of whether or not something is good.
    • mrweasel4 hours ago
      Imaging if people returned all the defective junk they got from Temu or Amazon. If the price is low enough people aren't returning anything, it's simply not worth the hassle and they'll accept that they got swindled.

      For books I'm not sure, assuming that the author actually have a story to tell, but is just lacking the skills to write a book, I don't know that I think you can justify calling it a grift when they're trying to write that book using AI. 370 books however.... Is she really that imaginative?

      There's a hell of a difference between having an AI help you write a fictional story that you've planned out and dreamed about for years, but just didn't have the skills to write and then pumping out 20 books a month. There's no way those were have been edited and read by one person.

      • skeledrew2 hours ago
        Some people really are that imaginative, and this author has been putting out successful works for many years before she started using AI.
  • trizoza5 hours ago
    The push to become 1000x engineer at big tech is nothing compared to the push to become a 1000x author at big publishing media houses :D
  • NorthSouthNorth4 hours ago
    There is a demand for these books AFAIK, they existed before AI. Short, easy to read and I assume write romance books designed for people that read them in a day.

    But now with AI they're just easy to mass produce I guess. What's funny is there were authors doing this with like gpt-4o and seemingly selling a bunch. I can't imagine they were very good but the people yearn for them.

    But 370 in just over a year and a half is crazy work lol. Also a lot of RSA is anti-ai so seeing a woman earning like 60 times the annual average selling slop must be upsetting.

  • wiether5 hours ago

      > She has sold more than one million books, by her own estimate, in digital, print and audio formats, earning more than $1 million so far this year.
    
      > The South African now heads an operation that has trained several hundred people in her method and is preparing to launch software for AI authors called Authorythm.
    
    So basically she is yet another grifter making money selling her method for "success"
    • prmph4 hours ago
      If she has made a million dollars, why does teaching her method make her a "grifter"?
      • basscomm4 hours ago
        Because the real money comes from selling the method to gullible marks who want to get rich quick
        • skeledrew3 hours ago
          I think you should give the article a - nother? - read. Books are selling, and she's teaching her method of how to generate books that sell to others.
      • nullsanity4 hours ago
        [dead]
  • thewhitetulip4 hours ago
    I have a very simple ideology here: if you don't bother to write your book, I won't bother to read it
  • josefritzisherean hour ago
    [dead]
  • prmph4 hours ago
    I don't know why people are so upset.

    She has not broken any laws nor done anything scummy; what she has done is studied the system and hacked it to her ends: a true hacker in the HN sense of the word.

    Now, I hate AI slop as much as anyone can hate it, but if society is so insensitive to actual quality that it will reward slop to the tune of a million dollars, the problem lies with that, not her.

    And to her credit, it is not like she is one-shotting books with a single prompt: she has a structured process starting from a human outline and ideas, plus human review at the end.

    • ChrisMarshallNY4 hours ago
      That type of workflow seems to be the "secret sauce" for successful LLM management.

      A lot of coders are using almost exactly the same approach, and it seems to be bearing fruit.

    • HolyLampshade4 hours ago
      > Now, I hate AI slop as much as anyone can hate it, but if society is so insensitive to actual quality that it will reward slop to the tune of a millions dollars, the the problem lies with that, not her.

      Very true, but perhaps beyond my rapidly expanding misanthropy I can also dislike the process that feeds it. I don’t see it any differently than people disliking companies that push unhealthy and addictive foods, or drug pushers, or any other potential societal malignancy.

      Just because we can make the thing and people will buy it doesn’t mean we should make the thing. And the more people that adopt the “lowest common denominator” production process the more we rob ourselves of the joy of creation itself.

    • soon_to_be_sb4 hours ago
      >society is so insensitive to actual quality that it will reward slop to the tune of a millions dollars, the the problem lies with that, not her.

      If the society is so dumb that it does not deserve to survive, then so be it.

      Are you up for that line of thought?

    • mrweasel4 hours ago
      Same reason why I'm upset when Temu ships a garbage plastic gadget. Someone must have wanted it, they ordered it, but I doubt that they wanted the quality that showed up. We've all seen the videos of what people think they order vs. what actually shows up. I don't get why you'd go through the trouble of wasting the resources on a scam product, rather than just outright robbing people.
      • tsaixingwei4 hours ago
        What if the Temu plastic gadget that you ordered actually turns out to be actually quite good and does what you’d expect it too? I think that’s the real comparison here instead of automatically assuming her books are crap that no one enjoys. Obviously enough people enjoy her books (some of it anyway) to make her money.
        • mrweasel3 hours ago
          That's absolutely a possibility, we simply don't know that, but they may very well be enjoyed by readers.

          People aren't returning the bad Temu products, it's not worth the trouble, could be the same with these books, assuming they are cheap enough.

          • skeledrew2 hours ago
            Are there reviews though? Many readers are always ready to publish reviews of stuff they read, so it's mostly just a matter of finding reviews of works by this author.
    • philipwhiuk4 hours ago
      > a true hacker in the HN sense of the word.

      First, the societal view on HN-approved hackers is not always positive.

      Second, society is bad at recognising systemic problems, that doesn't mean they aren't problems. Something with short term gain for long term harm is a particular problem for our society to handle.

      > It is not like she is one-shotting books with a single prompt

      It's not clear where she gets the original concept - maybe it's an original idea - fine. But maybe that's another few prompts. Which would make her little more than a meat proxy.

      > she has a structured process starting from a human outline and ideas

      It's not massively complicated though is it.

      > plus human review at the end.

      Yeah, it's the proofreaders having to work through hundreds of pages of slop I'm bothered about. Even if you assume the latest models are awesome it's not like this is Astra/frontier output is it?

      • viraptor4 hours ago
        > It's not massively complicated though is it.

        Complicated enough that very few people have award winning stories. Not even books - you can write 5 great pages and get relatively well known. But apparently it's not that easy.

    • brador4 hours ago
      “Human review at the end” is bs. No one is doing that for 370 books. I gave up at 2.
  • staleBot_tom1pr5 hours ago
    You need to be the kind of character who wants to see more cheap stuff in the world to pull this kind of shit. AI augmented troll hackers of civilians and small honest SMEs, invasive stalkers, self-proclaimed hard pushing media jerks bribing doctors and teachers and civilians, medical malpractice proponents, genAI augmented trolls...

    Too many people, once they understand that pseudo-tricky or smart folks can freely abuse stupid or naive or temporarily "crippled" people, go on to do exactly that.

    It just lowers all maxima, consequentially lowering quality of output and quality of evolution in those buckets, including the quality of rare events.

    Few people care about those very special tears of disillusionment and guilt they will shed on their death bed.

    Drug-induced forced laughter and pseudo-dominance over sabotaged and vulnerable reign superior in minds that fully outsourced "consciousness" to certain value systems.

    Ask yourself: when do non-functioning people on the spectrum, or people suffering from later stage Dementia and Alzheimer's et al feel most "conscious" ... that's a consciousness you can't outsource.

    • skeledrew3 hours ago
      > more cheap stuff

      What exactly is the matter with cheap stuff? Isn't it every buyer's dream to get the stuff they want good, fast and cheap?

    • ign0rant4 hours ago
      > Few people care about those very special tears of disillusionment and guilt they will shed on their death bed.

      That was beautiful. Thanks