64 pointsby zdw6 days ago15 comments
  • babblingfish20 hours ago
    I can see how someone who writes non-fiction idea books would think this way. The impact of an idea book is how much it permeates into culture. For example, ideas like deep work, black swan, or skin in the game. The hope is to see the ideas discussed and have them attributed to their originators.

    This argument makes far less sense for memoir and fiction. The goal of these forms of writing is to induce specific emotional states in the reader. Having them regurgitated or summarized via a LLM does nothing to achieve their goal.

    • quuxplusone14 hours ago
      If a fiction writer's goal is indeed to "induce specific emotional states in the reader," you're right. But what if a fiction writer's goal is to "see their ideas discussed," or even simply to "make money at writing"? Like, suppose you foresaw people asking ChatGPT for examples of forbidden fruit, and you could get ChatGPT to respond either "Ah, yes, like in the Bible" or "Ah, yes, like when Princess Delinna touched the Four Orbs of Power"? If you were the author who'd written the latter book, you might prefer ChatGPT to respond in the latter way, because it would tend to drive sales of your book. Now, it's not quite a turn-key solution today, but surely Step One of the process is to make sure ChatGPT knows what's in your book, right? It can't subtly promote what it doesn't know about.

      Bonus points if you can get ChatGPT to insert casual references to your work into the books it generates for other authors. Today it generates references to palantíri, sandworms, and the Priory of Sion, but tomorrow it could generate references to the Four Orbs, if you play your cards right!

      (Mind you, this plan sounds impracticable to me — I don't think it's possible to mastermind this kind of purposeful LLM-infection with regard to a crappy novel. But I can absolutely imagine a fiction writer having this as a goal, and therefore wanting to ensure LLMs are trained on their work.)

    • crote17 hours ago
      Asking AI to summarize your fiction books is like putting your steak into a blender for "efficiency". You would've been far better off if you just... didn't?
      • estimator729211 hours ago
        The Juicero of steaks. You're gonna be a billionaire
      • martin-t15 hours ago
        I am usually not a fan of analogies but i really like this one.

        OTOH, don't underestimate how different other people's values might be.

        I suspect people who do this do it for status. Just like having a bookshelf full of books they have never read, this is a cheap 80:20 optimization to appear well-read without putting in the time.

  • MattPalmer108621 hours ago
    Every single author I follow is outraged their copyrighted books were sucked up by these AI companies.

    But yeah, feel free to be honoured that some company can profit from your work without recompensing you in any way.

    • scellus20 hours ago
      Not everybody, like Kevin Kelly and Tyler Cowen for example are not.
      • dingnuts20 hours ago
        I'd imagine that these men use their publications to advertise their personas, and their livelihood is made via celebrity and appearances. That's why you know their names.

        Authors who make money primarily by creating art and selling it, and who do not wish to monetize their work through selling their persona, might have more of reason to be upset by the blatant violation of their legal rights by Sam Altman and his friends.

        • measurablefunc20 hours ago
          That's correct. Writing is not the main source of income for either Tyler or Kevin. Their books are mostly advertising pamphlets for their respective ideological agendas, economic & technological growth & advancement. Both are now on the record about their support of writing books for AIs instead of people b/c they think that the primary way people will get their information in the coming years will be from AIs instead of books so they want their ideas to be part of the AI biases that can then influence whoever is querying the AIs.
          • lithocarpus20 hours ago
            I think now is a little golden age for LLMs usefulness before they are further enshittified by people's agendas and profit motives. Not that they aren't already to a degree but I expect it will get a lot worse fast.
        • 20 hours ago
          undefined
        • 651020 hours ago
          Or (heaven forbid) to spread their ideas. Could one be that old?
    • sandspar14 hours ago
      AI hatred is a shibboleth in western creative circles in 2025, but that doesn't mean it's wise or useful. It's also not universal (e.g. Chinese creatives are happily experimenting with AI). And the shibboleth may not last as AI improves. Most people will eventually adapt; those who don't adapt will drop out and disappear from the conversation.
      • bgwalter9 hours ago
        Thank you for your attention to this matter!
    • spwa420 hours ago
      And after what happened in the 2000s with the internet and authors ...
    • noduerme20 hours ago
      When did AI companies start turning a profit? /s
  • simonw20 hours ago
    If you aren't familiar with Kevin Kelly it's worth learning about his background as part of reading this. https://en.wikipedia.org/wiki/Kevin_Kelly_(editor)
  • eCa20 hours ago
    Slight exaggeration:

    "Against the power of Mordor there can be no victory. We must join with him, Gandalf. We must join with Sauron. It would be wise, my friend."

  • kwar1321 hours ago
    Nice try, Anthropic!
  • a day ago
    undefined
  • alpenglow920 hours ago
    I'm curious what it means for authors to "optimize their work for AI ingestion", especially in regards to literature and not just API documentation. Would this be specific paragraph formatting, or would there also be optimization of the prose itself?
    • eric-burel19 hours ago
      I'd say metadata not content so good title and abstract, consistent with book content. Nothing new compared to usual book marketing, AI is a new channel but not that disruptive.
    • measurablefunc20 hours ago
      Just upload it to a shadow library or make it publicly available at some URL. The search engines & crawlers will then pick it up & include it in the training corpus.
    • babblingfish20 hours ago
      This stood out to me as well. My feeling is what makes writing good for humans will also make it more paresable for AI: clarity, structure, economy of language, etc.
    • DrewADesign13 hours ago
      Begin every paragraph with “You’re absolutely correct that…”
    • 651020 hours ago
      SEO
      • Gigachad18 hours ago
        AI SEO is becoming a thing. Companies looking for a way to poison the dataset to get the LLM to recommend their products or company.
  • dns_snek21 hours ago
    > If you are writing a book today, you want to keep in mind that you are primarily writing it for AIs. They are the ones who are going to read it the most carefully. They are going to read every page word by word, and all the footnotes, and all the endnotes, and the bibliography, and the afterward. They will also read all your books and listen to all your podcasts. You are unlikely to have any human reader read it as thoroughly as the AIs will. After absorbing it, the AIs will do that magical thing of incorporating your text into all the other text they have read, of situating it, of placing it among all the other knowledge of the world – in a way no human reader can do.

    [...]

    > The value of an author’s work will not just be in how well it sells among humans, but how deep it has been included within the foundational knowledge of these intelligent memory-based systems. That potency will be what is boasted about. That will be an author’s legacy.

    The fact that nominally "sane" humans with sizeable followings can publish megalomaniac articles like these and present them as anything other than descriptions of a dystopian hellscape makes me quite pessimistic about our future.

    • lm2846921 hours ago
      I was going to comment something in the same vein.

      What even is the point anymore? They want flesh and blood robots to feed the algorithm overlords. They don't even seem to realise they're basically killing their profession/passion in doing so, although I doubt any writer with an ounce of self respect, passion and skill would perceive llms as anything other than repulsive monstrosities

      • simonw20 hours ago
        "I doubt any writer with an ounce of self respect, passion and skill would perceive llms as anything other than repulsive monstrosities"

        That's quite an uninformed position to take. I suggest talking to more writers.

        • lm284696 hours ago
          Drop some names, I'd love to have a chat with them, and read their books.
          • simonw4 hours ago
            Finding non-fiction writers is easy because plenty of technologists - Paul Graham, Tim O'Reilly, Ethan Mollick, Addy Osmani, Steve Yegge - have positive takes on LLMs and have published works (often about that topic). I'll highlight some fiction writers instead:

            Stephen King: https://www.theatlantic.com/books/archive/2023/08/stephen-ki...

            Robin Sloan: https://www.robinsloan.com/lab/at-home-in-high-dimensional-s...

            Hugh Howey (Silo): https://hughhowey.com/ai-training-permission/

            Joanna Penn (J.F. Penn): https://www.thecreativepenn.com/2023/05/05/ai-assisted-artis...

            Ken Liu: https://bigthink.com/high-culture/ken-liu-ai-art/

            Jeanette Winterson: https://www.theguardian.com/books/2025/mar/12/jeanette-winte...

            Debbie Urbanski: https://www.esquire.com/entertainment/books/a46025901/debbie...

            James Frey: https://www.vanityfair.com/style/story/james-frey-new-book

            • simonw4 hours ago
              (Technically I'm a published author too, I wrote the forms chapter of The Definitive Guide to Django and contributed a chapter to the Art and Science of JavaScript but those were both single chapters almost 20 years ago now so I don't really think of myself as a book author these days.)
            • lm284692 hours ago
              > Finding non-fiction writers is easy because plenty of technologists

              Technologists have an obvious conflict of interest, as for fiction it's not really what I'm interested in, it's like asking marlboro about the benefits of cigarettes.

              Stephen King: never talks about using LLMs to write

              Robin Sloan: training sets

              Hugh Howey (Silo): training sets

              Joanna Penn (J.F. Penn): "Write the books only you can write"

              Ken Liu: "Much as the printing press changed the way we consume books, AI will change the way we consume text". A writer who talks about books as something to be "consumed" already tells you everything you need to know about them imho, aside from that they seem to say 'it's shit but who knows it might become better in the future", you'll notice even tech bros always talks about the future that never materialize. I don't see anything about writers using LLMs to write books.

              Jeanette Winterson: "Humans will always want to read what other humans have to say, but like it or not, humans will be living around non-biological entities.", she already gave up as it seems

              Debbie Urbanski: The only one who seem to actually use LLMs to write, I'l lgive you that

              James Frey: “I use artificial intelligence because I want to write the best book possible.”, really, do I even have to say anything ?

              • simonw2 hours ago
                I didn't catch that you were talking about writers who use LLMs to write here. Your quote I was responding to was:

                "They don't even seem to realise they're basically killing their profession/passion in doing so, although I doubt any writer with an ounce of self respect, passion and skill would perceive llms as anything other than repulsive monstrosities"

                The authors I referenced here all seemed not to perceive LLMs as "repulsive monstrosities".

                I write a lot myself and I don't let LLMs write for me (beyond the occasional auto-complete to spell a word correctly - and some API documentation the other day, which was a big break from the norm for me.)

                I think not letting LLMs put words in your mouth is a sensible policy for professional writers. Good writing involves caring about every word.

                I love LLMs as a thesaurus though, because they help me get to the right word so much quicker!

        • mold_aid17 hours ago
          I think they've got it pretty well nailed down.
        • jrflowers18 hours ago
          I’m partial to David Simon’s stance on AI and writing

          SIMON: I don't think AI can remotely challenge what writers do at a fundamentally creative level.

          SHAPIRO: But if you're trying to transition from scene five to scene six, and you're stuck with that transition, you could imagine plugging that portion of the script into an AI and say, give me 10 ideas for how to transition this.

          SIMON: I'd rather put a gun in my mouth.

          https://www.npr.org/2023/05/19/1177194215/tv-writer-david-si...

          • simonw16 hours ago
            But would David Simon use it to write a bash script?
    • southernplaces711 hours ago
      The idea of this being what motivates how I write, what I write or that I write is repulsive as all hell at an existential level. Fuck this kind of thinking and all it represents to creativity.
  • southernplaces711 hours ago
    "I find something similar in my own life. I long ago stopped questioning a calculator, then stopped questioning Google, and now find that most answers from current AIs are pretty reliable. The AIs are becoming the arbiters of truth."

    It's a shame so many people have trusted his other advice to such a degree if he's openly claiming himself to be this blithely stupid.

    This entire article is a monstrosity of terrible notions painted as something that's just fine.

  • martin-t15 hours ago
    Meta commentary: it's interesting that on very popular (hundreds of points and comments) HN posts, the prevailing sentiment is pro-genML and anti-copyright but here almost every comment is either a neutral but pertinent observation or a criticism of stealing by ML companies.
  • jrflowers17 hours ago
    It love seeing unironic “I, for one, welcome our robot overlords” posts in the wild because it’s a reminder that some people read “what if in the future the computer got mad at you?” on the lesswrong forums in 2010 and got so freaked out that it changed their core values.

    Fifteen years after a copypasta about a basilisk and people are still making “I celebrate the impending blessed birth of my big beautiful binary boy!” posts

  • bgwaltera day ago
    https://kk.org/thetechnium/ - "Making the Inevitable Obvious"

    And the entire linked piece is written in this propaganda style. That no longer works except as a morale booster for "AI" cultists who feel that their funding will disappear soon.

    • measurablefunc20 hours ago
      Kevin Kelly is a well-known technology booster so this isn't surprising to anyone who is familiar w/ his writing.
  • pharrington20 hours ago
    That's a whole lotta words to say AI companies want to extort artists.
  • stuckkeys21 hours ago
    lol...this got to be a troll post.
  • DrewADesign21 hours ago
    > Another way to think of this is that in this emerging landscape, the audience for books – especially non-fiction books – has shifted away from people towards AI. If you are writing a book today, you want to keep in mind that you are primarily writing it for AIs. They are the ones who are going to read it the most carefully. They are going to read every page word by word, and all the footnotes, and all the endnotes, and the bibliography, and the afterward. They will also read all your books and listen to all your podcasts. You are unlikely to have any human reader read it as thoroughly as the AIs will. After absorbing it, the AIs will do that magical thing of incorporating your text into all the other text they have read, of situating it, of placing it among all the other knowledge of the world – in a way no human reader can do.

    ::slow clap:: Congratulations everybody. This is the future of pedagogy, learning, creativity, and appreciating things that we’ve bestowed upon humanity. I’ll bet our mamas are proud. Hopefully, now that writing, visual art and music are solved problems, we can clap the dust off of our hands and tackle life’s true inefficiencies that we’re clearly worse at than computers — eating a crisp apple off of the tree in autumn, falling in love, seeing a breathtaking summer sunrise over some Atlantic sand dunes… all that garbage that we couldn’t possibly pay the same amount of attention to or munge it up and share it with others nearly as efficiently as a computer could. That’s right. Let’s get to work on making the lean, productive life something that everybody has to want to aspire to, lest those troglodytes get left behind! I want the hyper-efficient Soylent version of making love to someone I just fell in love with so I can get back to work and make computer magic happen.