How Are GPTs Affecting Book Publishing?
Advisor: Michael Festo
As an aspiring writer, I am interested in learning more about the publishing industry. Like most white-collar fields, publishing is undoubtedly changing with GPTs’ proliferation. This project aims to synthesize a comprehensive understanding of how a book is traditionally published and how the process is changing and may continue to change because of AI.
My principal research question is: How has book publishing changed after the proliferation of Generative Pre-trained Transformers (GPTs)?
I intend to conduct an extensive literature review and interview industry professionals, including authors, editors, category managers, and graphic artists, to learn about the publishing process firsthand and acquire different perspectives on how AI currently influences and will continue to influence the traditional publishing process.
Using the gathered research, I will craft a multimodal document (pictures, words, graphs) describing how aspiring authors can navigate the writing, editing, publishing, and promotion of human-authored work. This document will provide authors with various pathways to publication, as well as ways in which AI may harm and benefit them. The document may be shared with publishing houses, writers' associations, and technology companies involved in AI development for the publishing sector, contributing to ongoing discussions and future strategies in this rapidly evolving field.
Project Blog
Post 1: How Is AI Impacting Book Publishing?
As an aspiring writer, I am interested in learning more about the publishing industry. Like most white-collar fields, publishing is undoubtedly changing with GPTs’ proliferation. From my research, I intend to gain a comprehensive understanding of how a book is traditionally published and how the process is changing and may continue to change because of AI.
My principal research question is: How has book publishing changed after the proliferation of Generative Pre-trained Transformers (GPTs)?
I intend to conduct an extensive literature review and interview industry professionals, including authors, editors, category managers, and graphic artists, to learn about the publishing process firsthand and acquire different perspectives on how AI is influencing and will continue to influence the traditional publishing process.
Using my research, I will craft a multimodal document (pictures, words, graphs) describing how aspiring authors can navigate the writing, editing, publishing, and promotion of human-authored work. This Author’s AI Handbook will provide authors with various pathways to publication, as well as ways in which AI may harm and benefit them. The document may be shared with publishing houses, writers' associations, and technology companies involved in AI development for the publishing sector, contributing to ongoing discussions and future strategies in this rapidly evolving field.
My Author’s AI Handbook could be shared with publishing houses, writers’ associations, and technology companies involved in AI development for the publishing sector, contributing to ongoing discussions and future strategies in this rapidly evolving field and helping bridge the gap between authors and publishers.
Linked here is my full proposal.
Post 2: Week One: Preliminary Research
I spent this week researching self-publishing to decide whether I want to focus on it contra traditional publishing, which I do.
I began by reading David Gaughran’s “Ten Steps for Self-Publishing,” which helped me better understand the basic processes involved in successfully creating and publishing a book independently. I also purchased Gaughran’s Let’s Get Digital, the first of four self-publishing guidebooks, which I intend to read for my literature review.
To inform myself about industry trends, I also read “Self-Publishing Is Thriving” by Jim Milliot in Publishers Weekly, from which I learned statistics substantiating self-publishing’s increasing popularity and viability as an alternative to traditional publishing. Similarly, I signed up for Jane Friedman’s newsletter, a reputable source of self-publishing knowledge.
Joanna Penn’s Creative Penn project has also been another valuable resource for learning about self-publishing. She has hundreds of podcasts with self-publishing luminaries in which they discuss various informative topics; I have so far watched her “Self-Publishing With Print on Demand” video, in which she recommends print-on-demand services like CreateSpace and Lightning Source and outsourcing graphic design tasks like designing the book's cover.
My current plan for the near future is to research self-publishing’s genesis to provide my deliverable and final presentation with background information before examining AI’s influence and usage. I will begin working on my deliverable using the aforementioned research once I have completed it.
I will also begin scheduling interviews with self-published authors and several people working in traditional publishing to gain more background information for my deliverable
Post 3: Week Two: Researching Self-publishing's History
This week, I primarily researched the evolution of self-publishing.
I investigated the pre-ebook self-publishing community in California beginning with zines in the 1960s, documenting key findings and sources. I learned of the significance of the Espresso Book Machine, a relatively compact and portable book printing machine, the proliferation of which made Print on Demand services more accessible, which were integral to self-publishing before ebooks and are still important today.
I then researched the release of the Kindle and its impact upon self-publishing. I learned about how it made self-publishing significantly more accessible and financially viable for authors by providing them with an integrated ecosystem, effectively mimicking the experience of reading an actual book, and negating the need for printing of any kind.
I then began constructing a timeline outlining self-publishing’s evolution and role in the broader publishing market for my deliverable, on which I have begun preliminary work. I have also begun reading David Gaughran’s Let’s Get Digital, a well-regarded introduction to self-publishing.
Toward the end of the week, I finalized general interview questions and an outreach email for potential sources, sending them to a contact of mine at Barnes & Noble to disseminate and several potential interviewees sourced by my mom. I also revised my abstract to sharpen its focus and reviewed and updated my proposal’s bibliography per my recent research.
Post 4: Week Three: Conducting Interviews
This week, I primarily focused on scheduling and conducting interviews with self- and traditional-publishing professionals.
On Monday, I interviewed David Berkowitz, a marketing consultant and founder of the Serial Marketers community. He emphasized AI’s potential to democratize the publishing process, serving as a subsequent evolution to POD self-publishing and ebooks. He also discussed AI’s (hopefully) temporary drawbacks of market oversaturation and ethical quandaries regarding its plagiaristic nature.
On Thursday, I interviewed Edward Ash-Milby, an executive editor at Weldon Owen, an imprint of Insight Editions. He has contrastingly adopted AI little into his workflow, feeling it can’t replicate a human’s ability to edit and market books. He feels the understanding he gains from reading a book, with which he ideates edits and book descriptions, is necessary for his work and cannot be replicated by AI. He however sees AI’s utility in simpler tasks like indexing and metadata tagging.
Both interviewees feel AI literacy is integral for navigating modern publishing and that AI’s influence in the publishing field will only continue to increase.
On Friday, I have two more interviews scheduled with publishing professionals. Mr. Greg Boguslawski, the executive vice president at Blackstone Publishing, will hopefully inform me about another significant aspect of publishing for most self-published authors: audiobooks. From Ms. Tyrrell Mahoney, I am hoping to learn more about independent publishing, as she is the president of Chronicle Books, a publisher whose focus on smaller authors offers a middle-ground between self- and traditional publishing.
Post 5: Week Four: Interviews and Deliverable Work
This week, I primarily focused on conducting the last of my preliminary interviews and working on my AI guide deliverable
I interviewed Marty Schoenleber, a publishing strategist at IngramSpark, who discussed how AI is making marketing and metadata management more efficient, which is especially helpful for self-published authors seeking to compete with traditional publishers who offer more resources for delegating such tasks. However, Marty noted the potential homogenization of content and suppression of diverse human voices that comes with profuse AI utilization in publishing.
I also talked to Greg Boguslawski, the executive vice president at Blackstone Publishing, who highlighted audiobooks’ importance as a marketing tool and revenue stream for self-published authors. Mr. Boguslawski believes AI’s functionality and usage to create audiobooks is rapidly increasing, democratizing their creation by allowing self-published authors to independently produce them.
Tyrrell Mahoney, president of Chronicle Books, shared my other interviewees’ view that AI is useful for simple tasks like metadata creation, but she emphasized AI’s fundamentally plagiaristic nature and that she does not believe it should be used for any form of creative work.
With all of my preliminary interviews now conducted, I moved on to working on my deliverable toward the end of the week. I have completed my introduction and am working on the main guide section, aiming to have a draft done in a couple of weeks.
Post 6: Week Five: Deliverable Work and Research
This week, I primarily worked on my AI-in-self-publishing guide deliverable and conducted research on several key topics.
I found several more self-publishing luminaries I can contact for interviews, including Joanna Penn, Jane Friedman, and Mark Coker. I have also written a half-page summary for each interview I have conducted so far, noting three key quotes from each.
Notably, Tyrrell Mahoney highlighted that AI tools can “provide efficiencies,” yet they also present ethical quandaries regarding plagiarism and intellectual property protection. Greg Boguslawski emphasized that while human work is still integral to audiobook creation, AI makes certain tasks like editing take “minutes rather than real time of like 10 hours,” and its relative accessibility has been steadily democratizing audiobook creation.
I have also been researching interesting AI tools for self-publishing. Sudowrite, Quarkle, and Elevenlabs have stood out to me so far as saliently useful AIs for ideation, editing, and audiobook creation, respectively.
In the coming weeks, I will continue my research into these AI tools and hopefully schedule interviews with more self-publishing professionals. I aim to have a rough draft of my guide deliverable completed for feedback from my in-school and external advisors within the next couple of weeks (I was originally planning on sooner, but underestimated the time college visits would take and the quantity of information I have amassed to parse).
Post 7: Week Six: Research and Deliverable Work
This week, although unable to meet with my project mentor, I made good progress on my independent research and deliverable.
I have been researching candidates for further interviews; I have three strong ones so far: Joanna Penn, Jane Friedman, and Mark Coker. These eminent self-publishing professionals have extensive experience with AI and self-publishing, generally supporting its integration. I believe they offer differing perspectives, however, in their experiences with AI, as Joanna Penn has direct authorial experience with it, Mark Coker’s understanding is likely more of how AI is impacting the self-publishing market holistically, and Jane Friedman can speak to some of AI’s downsides given her personal experience with being a victim of Intellectual Property theft facilitated by AI.
I have also been researching niche, interesting AI tools for aspiring self-published authors. Sudowrite, Marlowe AI, and ElevenLabs have interested me most so far. I have found Sudowrite effective at extending preexisting writing and generating logical plot developments (this is just with the free version; I have yet to work with the paid one). Marlowe AI provided me with a number of statistics regarding the text I submitted, including what percentage was dialogue vs narrative and the quantity of cliches, repetitive phrases, and passive voice usage (the paid version offers more detailed analysis regarding plot structure, comps, story beats, etc. which I haven’t had the opportunity to test yet). ElevenLabs’s AI voices are very realistic and effective; paired with a more versatile AI tool like ChatGPT, I believe one can efficiently generate and proofread AI book narration for their work (this is also something I will test).
Assuming my continued testing finds these AIs suitable for incorporation into my AI Guide, I will describe how Sudowrite can be used for the early stages of ideation and writing, Marlowe AI can be used for initial proofreading and editing (before hiring a professional human editor), and ElevenLabs can be used for creating audiobooks in numerous languages for increased marketing and salability.
I have also been working on the final section of the introduction for my AI Guide. This portion synthesizes information from my previously completed interview transcripts and summaries, as well as findings from my literature review. When finished, the section should provide readers with a clear, accessible overview of AI’s growing significance and general use cases in self-publishing, coming directly from reputable specialists.
I aim to have a full draft of the intro section completed next week and a draft of the main section done the following week for feedback from my mentor and in-school advisor. As these sections are completed, I can provide more details regarding their content and structure.
Post 8: Week Seven: Finalizing Interview Candidates and Finishing Deliverable Intro
This week, I updated my AI tools list, completed a rough draft of my deliverable’s introduction, and researched three more candidates for my final set of interviews.
I will send my deliverable’s intro section to my faculty advisor, Mr. Festo, and my outside mentor, Thad McIlroy, for feedback. The intro section contains an abstract, a detailed introduction distilling the key themes that emerged from my completed interviews and literature review, and a timeline of self-publishing’s recent history. It is designed to introduce the reader to self-publishing, AI’s general uses and misuses within self-publishing, and what my guide will offer.
I also refined my shortlist of exemplar AI tools. Per my mentor’s suggestion, I’ve replaced Marlowe AI with Quantifiction, whose manuscript reports draw on newer large-language-model architectures and offer a more comprehensive and detailed analysis, including tonal shifts, reading-level heat maps, and character-centric sentiment arcs. This wealth of features should allow self-published authors to refine their manuscript to the extent that minimal work with a human editor is required, saving time and money.
Finally, I researched three additional interview candidates my mentor identified who can contribute to my guide.
David Wogahn: President and founder of AuthorImprints; 40-year career in media and publishing technology; helped more than 120 self-publishing authors and businesses publish over 350 books using their own imprints; author of five books and a recognized consultant, known for his expertise in professional self-publishing, ISBNs, imprints, and book marketing strategy. He will add a more business-oriented perspective to my project.
Dan Holloway: Self-Publishing News Editor at the Alliance of Independent Authors (ALLi); produces ALLi’s weekly Self-Publishing News podcast, offering analysis of significant developments in self-publishing, including regulation, copyright, and technology shifts; recognized for his broad, policy-focused view of the publishing landscape, informed by his background in both creative and journalistic work. He will add a more policy-oriented, industry-wide perspective.
Matty Dalrymple: Author of The Ann Kinnear Suspense Novels and Lizzy Ballard Thrillers, as well as non-fiction books on the writing craft and publishing business; hosts The Indy Author Podcast and serves as Campaigns Manager at ALLi, where she supports independent authors through advocacy and education; regularly speaks and consults on writing, publishing, and platform-building, offering practical guidance from her dual experience as a working author and industry advocate. Provides a balanced perspective with a nuanced understanding of the business side of publishing as well as the craft side.
Next week I will reach out to my finalized interview candidates, integrate advisor feedback into the intro, and begin working on the deliverable’s main section to have a rough draft completed for review ASAP.
Post 9: Week Eight: Working on Deliverable Main Section
This week, I began my AI Guide deliverable’s main section and addressed feedback from my two advisors.
Having completed and circulated my AI Guide’s introduction, I began working on my guide’s main section. After discussing timelines with Mr. Festo and my outside mentor, Thad McIlroy, I determined that pursuing additional interviews is unnecessary and too time consuming. I believe I already have enough information to complete the main section of my deliverable, and with only a few weeks to finish it, I think I should focus entirely on completing it. Should I need more information, my outside mentor has numerous contacts with whom I could likely speak on short notice, and several of my previous interviewees offered to provide further assistance if I reach out to them again.
The feedback I received on my guide introduction from my advisors was very helpful. I have been working on implementing it, namely, after correcting smaller solecisms and formatting errors, incorporating my abstract into my introduction and adding more visual elements and statistics to supplement my initial surplus of interview anecdotes.
The main section, as stated in previous blogs, will be influenced by David Gaughran’s “How To Self-Publish A Book In 10 Steps.” For all steps to which AI applies, I will provide mild (diagnostic tools like Quantifiction or LLM ideation), moderate (collaborative drafting with tools like Sudowrite), and maximum (full LLM rewrites and novel writing, generating a cover with AI) workflows. This structure will let authors determine the extent to which they intend to incorporate AI, which is significant not only artistically and creatively but also for marketing, as some self-publishing platforms (e.g., Amazon) require direct AI use to be specified on the book listing.
I will also include a section dedicated to discussing audiobooks and AI’s facility to generate them, as they are significant marketing tools and revenue streams for many self-published authors.
From here, I am continuing to work on my deliverable, of which I will be sending drafts periodically to my two advisors for feedback, and continuing any necessary research this work requires.
Post 10: Week Nine: Deliverable Work
This week, I finished two more of the eight total sections of my AI guide deliverable.
My current plan for my AI guide is for it to have an intro section, including general information about self-publishing and AI’s use in it; a main section with six subsections detailing specific AI use at major self-publishing steps; and an outro section including a conclusion and future predictions about AI’s significance in self-publishing.
I have completed a draft of the intro section and the Writing and Editing portions of the guide's main section. I aim to finish the main section, which will include Writing, Editing, Branding, Design, Metadata, and Marketing, in the next week and a half. This will give me another week and a half to complete ancillary tasks, including revising the intro section, writing the outro, and adding a section, likely in the intro, detailing major self-publishing platforms' policies on AI-use disclosure and how the Minimal, Moderate, and Maximum Involvement categories fit into disclosure.
I am also still confident I have enough information and do not need to conduct more interviews, as the majority of my information is coming from my literature review, with the interviews serving more to establish ethos and offer the opinions of professionals that would otherwise be inaccessible. Most of the AI tips that comprise the guide are from my own experience working with the AI tools I discuss, as well as reputable sources like Kindlepreneur and Future Fiction Academy.
From here, I am continuing to work on my deliverable’s main section, of which I will send a draft when completed to my two advisors for feedback, and continuing any necessary research this work requires.
Post 11: Week Ten: Deliverable Progress
This week, I finished another two portions of my deliverable’s main section.
I first combined the Branding and Design portions, as I determined they are fundamentally inextricable, and AI is only really applicable to cover design; other aspects, like formatting print copies and ebooks, can be easily done with templates. I then completed this combined portion as well as the following Metadata portion.
I did not initially realize how significant metadata and branding are for self-publishing, learning while completing these subsections of the numerous ways in which one can increase their book’s salability. These include more obvious aspects like refining a book’s title and cover to factors I didn't initially consider like author bios, book descriptions, and keywords one can choose on a platform like Amazon. I have found numerous interesting uses of AI for these purposes, including researching successful books in one’s genre and finding patterns between their metadata, which I think are very good for self-published authors to know.
These updates mean I have now completed the Writing, Editing, Branding, and Metadata portions of my deliverable’s main section, with only the Marketing subsection left. I expect to finish the main section by early next week, leaving roughly one and a half weeks to revise the intro, complete the outro, and prepare my presentation.
I have decided to incorporate a disclaimer in the intro describing the policies of major self-publishing platforms regarding AI-use disclosure and how the Minimal, Moderate, and Maximum Involvement categories for each subsection fit into disclosure. The main section already aligns well with the policies I have been researching; I hope to have the Minimal category align with not needing to disclose, the Moderate be more up to the author depending upon their usage, and the Maximum definitely require disclosure.
From here, I will continue researching and working on my deliverable’s main section and will send a draft to my two advisors for feedback once completed.
Post 12: Week Eleven: Deliverable and Presentation
This week, I completed the main section of my AI Guide deliverable and started working on my presentation as my project nears completion.
I finalized the Metadata and Marketing subsections and sent them to my advisors for review. Their feedback was invaluable, providing insightful points I am addressing into the final week of the project. These insights are crucial for refining my work and ensuring it is effective in its goal to inform and aid self-published authors.
Additionally, I completed the AI Disclosure section of the introduction. This part discusses the policies regarding AI use and disclosure on major self-publishing platforms, such as Amazon KDP, Audible, and Apple Books. It provides an overview of how my Minimum, Moderate, and Maximum categories align with these disclosure policies. Minimum indicates that no disclosure is necessary, while Moderate signifies that disclosure is required. Maximum involvement suggests a more extreme use of AI, which should only be considered if one is low on time or money. This section aims to provide readers with a thorough understanding of the implications of AI use, ensuring they can navigate self-publishing platforms effectively.
From here, my primary tasks include making final revisions to both the intro and main section, writing the outro, and completing and practicing my presentation. I am confident I have the time and information needed to successfully complete my project on time.
Post 13: Week Twelve: Sign Off
This week, I completed, submitted, and presented my project.
I enjoyed having the opportunity to share my research with peers and faculty, and appreciate all the support I received from them. I am eager to see my advisor’s feedback on my final deliverable, which will undoubtedly benefit me in revising my AI guide deliverable for potential future publication.
My outside mentor, Thad McIlroy, has offered, provided I revise my guide to publishing standards, to publish it through his company, The Future of Publishing. I am very grateful for this opportunity to begin a hopefully long and successful career as a published author.
I am also grateful for everyone who has supported my blog by reading and commenting. I am happy to see my passion for writing and AI reflect in others. I have also really enjoyed keeping up with my peers’ research, learning about a variety of interesting topics I otherwise wouldn't have in the process.
From here, I will be continuing my research and revising my guide deliverable over the summer to hopefully have it published. I have really enjoyed learning more about publishing and AI, and am excited to continue to do so.