In this solo YouTube presentation, Julie Broad (founder of Book Launchers) explains why the recent academic study “Bought or Not?”—which found that readers rate AI‑written short stories higher in quality and cannot reliably identify AI authorship—does not mean nonfiction authors should hand their manuscripts to AI. She walks through the study’s three experiments, highlighting that AI stories scored higher on quality and absorption, but readers only identified AI authors correctly 39% of the time, and that attribution (telling readers a piece is human‑written) boosts perceived quality. Broad argues that the study’s short‑story context does not translate to 200‑page nonfiction, where authorial humanity, personal stories, and credibility are essential for authority, discoverability, and word‑of‑mouth promotion. She notes that many AI‑drafted manuscripts look polished but lack the “heartbeat” that connects readers, a gap editors cannot fix. She outlines three key takeaways: (1) clarity beats cleverness—AI is clearer but human writing can be overly complex; (2) the attribution bump—readers pay a premium for a human author; (3) AI familiarity is rising, so the window where readers won’t notice AI is closing. Broad offers practical advice: embed personal anecdotes, ensure the manuscript contains content only the author could have written, use AI only for outlining, clarity, and marketing, and leverage grassroots, local‑first promotion. She promotes Book Launchers’ free audio guide “Self‑Publish and Succeed” for authors who want to write books that readers will recommend.
Chapters
1,682 readers just rated AI-written stories better than human ones, and when they asked readers if a book was written by AI, they got it right less than 40% of the time. Before you hear that and hand your manuscript to ChatGPT thinking, okay, cool, let's write my book, there's also some more information from that study you need to know. So if you're a business owner, a speaker, an expert writing a nonfiction book to build your authority, well, This video is for you. Hi, I'm Julie the Book Broad, founder of Book Launchers and your host here at BookLaunchers.tv. Book Launchers' mission is to help 100,000 books be discovered by 2030. And book discoverability, it starts with positioning and writing a truly great book. And that is why this study matters so much. So what's the study? Well, let me give you the actual paper because if you've seen headlines, that may not be the exact thing we want to talk about. But the paper is called Bought or Not? by Sydney Sears and by Dina Skolnick-Weissberg. It was published August 5th in Judgment and Decision Making, which is a Cambridge University Press journal. So this is real research. They conducted 3 experiments. In the first one, 1,682 people read short stories. Some by humans, some by ChatGPT, and they did not know which. The AI stories scored higher on quality. They scored higher on absorption, which is a fancy way of saying I couldn't put it down. In the second and third, about 900 more people were asked to just pick which one the machine wrote. When they were asked to pick, they got it right 39% of the time. That's not even half on a 2-option question. Flipping a coin, you might have done better. So here's the part that should have you sitting up. The ones that knew a lot about literature, they didn't do any better. People who widely read or know fiction, they were not better at spotting AI or not. The only thing that predicted getting it right was how familiar someone was with AI itself. There's a couple things that my team has been talking about. One says, well, AI can't really write. Readers can tell and anyone with taste can tell. But the study really just refuted that. Readers can't tell, not reliably, and not at a better rate than chance. Now, the other camp of people said, great, readers can't tell, so AI can write the book. I'll save myself $20,000 and 6 months of pain. Both are reasonable-sounding arguments, but I don't think either are right because the researchers ran the same stories again, but this time they told people who wrote them, although sometimes they lied and sometimes they told the truth. And when readers were told a story was written by a human, they rated it higher, about 2/10 of a point higher on quality, but they still rated it higher. Same story, same sentences, same commas, but a higher score. When they were told it was AI, the rating dropped substantially. Again, same stories. Here's the thing from this study. Readers can't tell, but they pay for believing you are real. And now, okay, I want to make sure that we're clear. The study used short stories. These were fiction, not a 200-page nonfiction book about scaling a services business or leadership or financial prosperity. And I think that this distinction does matter more for you. A short story is a pure test of writing. It's words. Your book, it's not that. Your book has your name on it, your face on the back, your company, your clients, your reputation is all there in that book and somebody is reading it deciding whether to hire you. The attribution effect they measured in a controlled lab is not a small variable in your business, it's the entire thing. The fiction author is selling a story, entertainment, an escape, I don't really need to trust the author is real if I am getting entertained by something. But if I read your book on leadership or fat loss, well, yeah, I want to know that you know your stuff. And you as the author, I want your story in it. You sell the fact that you're the one telling it. And I want to tell you what we're seeing at Book Launchers right now, because it's changed so dramatically in the last year. We are getting so many manuscripts coming in saying they're ready for publication. Can you just put a cover on it and design it? It's edited. I'm happy with where it is. And we dive in and we read it and we go, yep, it's got good structure and it's grammatically correct. But it's not a good book. And the book might be fine. But fine doesn't create a great connection with a reader. And ultimately, The true success that comes from being an author comes after people read your book, recommend it, review it, they tell other people. And what we're finding is these books, they just lack that connection that's going to have people recommending the book. Now, they also often have subtle mistakes, which you should definitely check out the video I talked about where AI is putting lies in your manuscript. Now, I'm not throwing anyone under the bus because there are lots of good people with real expertise who tell the AI systems, their expertise, and it gets put into a book and it has all the right parts, but it's still missing that heartbeat that a reader can ultimately feel over 200 pages of a book. And this is not something an editor can fix. An editor can fix the grammar and the punctuation, but an editor, their job is not to install a heartbeat into the book. To do that, it requires rewriting. It requires going back and putting more humanity in line by line. And most authors have no idea that's the situation they're in because the draft in front of them looks finished and polished. And nobody's pricing that cost in when they say AI saved them 6 months. You're sitting there going, well, the study said AI stories were more absorbing. People got more stuck into it. Well, the study used short stories, a few minutes of reading. We're talking about hours of reading. And AI right now can absolutely win 5 minutes. It's clean, it's smooth, nothing really trips you up. But is it winning 200 pages? The study didn't look at that. And having looked at hundreds of manuscripts over the past year, I can tell you AI is not winning the reader yet on that volume. It lacks connection to keep you turning the pages over and over again. Absorption over 5 minutes and connection over a whole book, those are not the same thing. Now, the 3 things this study says that I think are worth paying attention to. One, clarity beats cleverness. And absolutely, I agree with that in a 5-minute story or in a 5-hour book. But here's why the AI stories won. The lead researcher, Weisberg, said AI writing is clearer and more direct and easier to process, and human writing is more subtle and complex. And she said people prefer that predictability because Difficult or subtle material requires more brainpower. The machine didn't out-imagine us. It didn't out-feel us. It was just more clear than most human writers are. It wrote in shorter sentences and got to the point faster, and it wasn't worried about sounding smart. We see this in so many nonfiction authors. They use big words because it makes them sound smarter, but it's actually way harder to read. You know, expert is brilliant, and they write a book using the big words that they've acquired. With all of their hard-earned knowledge. But complicated doesn't equal credible. We need clear. You gotta remember that most of the time when somebody's reading a book, they are winding down from a long day. It might be 9 PM, or they might be sitting there, you know, waiting for dinner and a kid's asking for a snack while they're trying to focus. If your paragraph makes them work too hard, they'll put the book down and not pick it back up. I think it's important to realize you're probably losing more readers to your own complexity than you are when they think you've written it with AI. Here's a test. Read your book out loud, and every place that you get bored or you run out of breath, that might be a sentence that you want to change. The attribution bump is the most useful number in the study though. The same words, when somebody was certain it was written by a human, they got a higher score. And that is a premium the reader is handing you just for being a human. And I think that's really important today to be sure that you're putting your name and your face and saying you wrote this book. This was your hard-earned sweat. And building a campaign while you're writing to show you writing the book, I think, can go a long way. The reader is not just consuming information because they can get information everywhere. They are really consuming your book because they want to connect with you They want to know what it was like to do the thing that you've done and know the person who's done it. So the parts of your book that can carry your fingerprints, that's really the asset that you have. Okay, one of our authors is doing this so amazingly, Tierra Duncan. You can follow her on social media and see, but she has written this book and she's working on her next book. She's put humanity into everything she does. She's been sharing it on social. And she's getting speaking engagements, media coverage, and she's building an audience with this book that's gonna be ready for her next book. And that happens because not only has she created a great book, but readers are connecting with the author that wrote it. And this is why grassroots marketing is working so well right now. If you start in your local community, people get to know you as a human, and they'll tell other people about it. And that will grow. Don't head straight for national media coverage. Focus on your local area and grow from there. And when you're looking through your book, if a passage could appear in any other author's book on your topic, cut it or make it yours. Add a story about your client experience, a time when something went so wrong. Be sure to share your personal stories. One thing you can do is go through your manuscript and count the moments that could have only come from you. The client who fired you, the launch that flopped, the number you got wrong and got embarrassed about, the Tuesday that everything went wrong or everything went right. And ultimately, make sure that you've got your stories, because if a passage could appear in any other author's book on your topic, cut it or make it yours. That alone will change your manuscript. Whether you used AI or not, that is a great way to approach writing a book, because AI could have written it if the book passage could have appeared in somebody else's. Now remember what predicted whether someone spotted the AI? Not literally knowledge. It was familiarity with AI. The general public is getting more and more familiar with AI. So this is something to pay very close attention to because every month more of your readers are using these tools every day, which means more often people are starting to recognize the rhythm and the patterns that are happening. That 3-part structure or the really short sentences. You see this in LinkedIn posts all the time. It's like the 2-word sentences. Right now, 39%. That's how many people recognize when something was written with AI. If your plan is readers won't notice, well, you're betting your reputation on a window that's closing. And you don't get to unpublish a book. It sits under your name permanently with reviews attached. I'm not anti-AI. I use it all the time. I use it to outline, to fix the clarity problem from point 1, to help generate book descriptions. And to ideate on marketing. But the moment it's generating the substance instead of sharpening it, well, you're trading the one thing readers measurably pay a premium for— your humanity. And if you're looking for some support to help you with this, you can go ahead and grab Self-Publish and Succeed to help you write a book that readers want to read. You can grab a free copy at booklaunchers.com/freeaudio. That will get you A study just told us that a machine can write a clearer sentence than most of us. That's probably true, and I'm okay with it. Let it fix your run-on sentence. That's totally fine because that was never really your advantage. The advantage is that you showed up, you did the work, you fell on your face and stood back up, and then you're telling the story after so others can learn from it. And then you put your name on the front and you're out there talking to people about it. Readers want to know who wrote it. Wrote the sentence, and they can absolutely tell if you had experience and expertise and your own failures baked into it. Ultimately, my advice to you is write the book only you can write, and then write it clearly enough that somebody actually gets to the end so they can recommend it to other people.
