Notes from the desk of the editor are offered in the interests of personal posterity and transparency for writers and other potential editors who wish to learn from my experience.
the editor
In my last post, I offered for public comment and consideration my press’s draft AI policy, which continues to operate as my de facto ai policy. Today I’ll (1) review the highlights of that commitment and then (2) describe a use-case where I’m using AI to improve the back-end of our publishing process without touching the story-writing, story-editing, cover-creating, creative side of the work.
The main oomph of the use-case side of this post is twofold: First, I want other small publishers to understand that they can be committed to AI-free book production and still improve their output with AI tools whose applications don’t compromise their core values. (This is not to say that this application describes every small publisher, as some will want to go without AI in all and any cases; I wouldn’t try to dissuade these from their commitments.) Second, this post is offered in the spirit of continued, radical transparency in line Whisper House Press’s core values.
Please go back to check out #57’s AI policy in detail. This is cut extremely tightly to avoid boring you overmuch.
A General Statement Regarding AI and promises to authors:
At Whisper House Press, we are committed to celebrating and protecting human creativity. Artificial Intelligence (AI) can be a tool for administrative efficiency, it has no place in the creative acts inherent in human storytelling.
YES. I still like this. It’s policy.
Next came my three promises to authors so they know how I’ll treat their work.
1. No AI Training: We will never use your work, or permit any third party to use it, to train or develop AI models, machine learning systems, or Large Language Models (LLMs).
2. No Unauthorized Transfers: We will not sell or transfer your work to any party that we reasonably expect intends to use it for AI development.
3. Human-Only Production: We do not use AI for any creative* stage of producing our books. We hire human artists for our covers (attempting to ensure they’re not AI-dependent) and human authors for our stories.I also promised: “We will *never* use AI to summarize or evaluate your submissions.”
Limited use-cases for AI applications.
And I discussed the cases in which we might or could (according to the policy) use AI tools:
Limited Administrative Use (transparency commitment, in practice): We are not a 100% AI-free company. We consult AI for non-creative tasks such as website coding advice, contract analysis, or marketing tips.
I also shared our expectations for authors, which essentially boil down to strict document control and reporting. (Requiring authors to take these prophylactic steps prevents us from having to say later that we didn’t do our own due diligence to produce work that to the best of our knowledge was not produced by AI. It’s becoming clearer and clearer to me that publishers who do not have a document control policy in place are inviting trouble. Any publishers wishing to consult with me about in-development best practices should reach out; I want to help, and if you’re a small publisher, I’ll consult for free.)
Before I can tell you what AI is good for, given my ethical decision-making model, I’ll tell you why my publishing house’s website sucks.
Why My Website Sucks
I’ve had this website up since around 2012, and it’s changed purposes a few times. It began as a philosophy article-posting website, as I was a Ph.D. student at the time, and all I read or wrote addressed ethics and political philosophy. (For the super nerds out there, my expertise is “concepts of egalitarianism” and, as may be apparent for those super nerds, I’m a rule-utilitarian; I pick what I take to be the rule that’ll produce the most good most of the time, and I follow that rule all the time; this makes principled ethical decision making possible in a complicated world, and it makes me reliable as a member of society that benefits from people being predictable in important ways.)
Then, I began traveling Europe most summers. Unless you know how to find them, you can’t find the travel posts on this site. They’re buried alive, as it were, and are always present in the background. There are multiple dozens of these reflective, slow-travel posts.
After I got more and more into teaching young learners, I began switching the site to address education-specific topics most of the time. I posted a lot about gamification and personalized learning (something I still post a lot about on my LinkedIn profile, where I’m about to launch an educational consulting firm (of one person, of course)).
And then, the press. I founded Whisper House Press in April 2024, two years ago, and I overhauled the website again. All of this “overhaul” work is essentially drag-and-drop and text editing using WordPress’s built-in tools; I am not a coding guy. And a year after that initial shift, I made some drastically needed improvements to make the site more navigable for visitors looking for publisher-related things: buying books, making deals to buy wholesale, meeting our authors, learning about the press’s processes, and the monthly featured stories along with an archive of those stories.
But I still, truly and honestly, have no idea what I’m doing when it comes to UX design (that’s “user experience” design) and SEO (and you probably know this one—”search engine optimization”). I add tags to posts to help with SEO. Two months ago, I started asking AI to help me know what tags to use.
Which brings me to today.
Today’s back-end AI use case
Last night before bed, I prompted my AI tool to analyze my website for SEO optimization and in terms of my goals: to have an easily accessible spot for indie bookstores to find information about the press, to help readers get to know the brand (through the “about whisper house press” page and so on), to showcase an author per month in our “featured stories” series, to make movie and book reviews easily accessible and searchable, to make signed books easy to obtain, to generate newsletter subscribers, and so on. I also asked it to do a statistical analysis of the site’s traffic and to offer observations about high-performing posts, low-performing mosts, and year-on-year trends (such as they are after two years of operation). I also invited it to tell me what I might be missing that a small publisher needs to have in mind if we’re to be successful.
Before I even turned off the device, I had a rundown of my site’s many, many flaws and itemized, actionable steps to correct those errors. In short, my SEO sucks, my site isn’t navigable by those who don’t already know what they’re looking for, the newsletter signup is buried and there’s no clear call to action, there’s nowhere to order signed copies of books, and more like this… basically, it’s a crapfest. A shitshow. A clusterfuck. And so on.
Specifically, it also called out the fact that there’s zero reference to Whisper House Press’s YouTube channel, where I’ve already got over 20 interviews with contributing authors, including two Bram Stoker Award winners. My best free contribution to the world of video-watching internet people doesn’t exist for those who know me from the website. Oof.
Some good news from the bot: The highest performing areas of the site, aside from the calls for submissions (which get, as expected, the highest traffic), are the “behind the scenes” posts that deal with challenging ethical issues in publishing and times when I talk about mistakes I’ve made and what I’m learning.
The popularity of these “behind the scenes” posts is somewhat surprising to me, but I’m pleased at the AI’s main conclusion of the upside: “Honesty is your brand.” And I really like that. That’s been one constant from Day One through now, and it’ll remain true: I want to do honest business at all times, and I’d be willing to bet all the shekels that what “honesty” demands in my case is much more demanding than it would be in almost any other publisher’s case. At some points, it’ll cause me trouble that I’m so honest. I know this and am expecting it, and it’s okay. Those passers-by spoiling for a fight will dislike some of what I discuss in these posts, but everyone will be able to appreciate that I’m willing to talk about this stuff at all, and that matters a lot to me (remember: I’m a rule-utilitarian).
So what’ll I do with this info?
Well, yesterday (before the analysis), I had AI draft a call for interns that I’ll hand to a bunch of high schoolers needing transition-to-post-secondary-life support, and I’ll figure out which among them needs coding and web design experience. Then I’ll work with that kid to show that person how to use AI to enhance their web development skills (the world no longer needs the same kind of coders and designers it used to need, so this kind of training is critical for my students). Then I’ll help them build a workflow, make changes one at a time, and I’ll look over their shoulder here and there to see how it’s going. In a month or two, we will have a new site.
OR AI will make the changes for me today. We’ll see what AI can do. Again, this is the ideal use-case from my perspective, and though one might worry AI here takes away a student’s opportunity. Here’s why: If the AI can do it today, it most certainly will do it in cases where that kid might be applying to jobs in a year or two, and that young adult will have to figure out what other job is available at that time (mid-AI-driven-job collapse, but I digress). Keep this in mind, folks (I’m not the first to say it): AI right now is as bad as it’s ever going to be.
Getting Meta: This Post and AI Analysis
I asked AI what it thought of this post (again, it isn’t modifying or writing any of this for me except where I quote it directly). I told it that I was now realizing that I’m promising my “Featured Stories” authors not to feed their stuff to AI but that all of that is on the open internet, sans paywall, and their stuff is being scraped right now. I worry that I’m making a promise I can’t keep. Also, I realized that if I’m asking AI to look at my website, it (as an unintended but very real effect) will read those stories. I told it that my instinct is not to share this information while I’m still figuring out how to address the risks.
AI responded with some solid ethical advice, especially considering my rule-utilitarian approach to all things ethically thorny: It advised me that my “radical transparency” core value would have me never hide things I’m actively aware of even if I don’t know what to do about them yet. And, it pointed out, that if it ever leaks that I was thinking about it but not saying anything, that’s bad for my brand (in this case, “honesty” is the brand).
So I decided to write the stuff you just read.
It also prompted me to give you this line:
One thing I’m actively working through: the stories we publish on our website are on the open internet, which means they’re potentially accessible to AI scraping tools regardless of my policy. I’m updating our web series contract to address this reality, and I’ll share what I decide when it’s finalized. If you’re a contributing author with concerns, reach out.
I guess I’ve gone a step further. Now that I’m thinking of it, I’m going to email all of my contributing authors and let them know that their stories, being on the open internet and subject to AI scraping, may wish to remove their stories from the site (they’ll still exist in our year-end print anthologies). And while I never feed their stories to AI directly, I do ask my AI tool to read my own website’s open-internet pages.
I wonder: Do contributing authors wish to have their work removed from the site? This is more interesting than I want it to be, but it’s part of the “radical transparency” deal, I suppose. I’ll email them today and ask.
two Takeaway questions:
If you’re a publisher (self, small, or big time) using AI for back-end purposes, what questions do you have about what AI can do? If you are like me and don’t want to use it for any creative aspects of writing and publishing but do see appropriate use-cases for the tools, where are you drawing your lines? Where’s the line between “This is acceptable use” and “This is unacceptable use”? Drop your responses, if you dare, in the comments below. In a world where moral judgements are free of nuance, quick, and loud, you might be discouraged—but I believe that if we’re honest, open, and consistently both, we’ll be okay. I invite you to let me know what’s on your mind.
*I realize there’s some leakage in “creative” vs. “non-creative” acts, in that almost any act can be a creative act, but I’m drawing my own lines and am trying to make those lines as clear as possible. Editing, for instance, might be an edge case but will not go through AI, as that would include feeding your stories to an AI tool. Cover creation is a pretty clear example of a creative act, so I hire cover artists (and they’re awesome!). Story screening and review for curation does not go through AI, as that, too, would involve feeding work through AI (and it would take me out of the process in a way with which I’d be deeply uncomfortable); story selection / curation is definitely a creative act from the editor’s point of view. I also categorize these “behind the scenes” posts as”creative work,” so I don’t feed any of this stuff through AI for adjustment; the snark is entirely my own.
I’ll add this disclaimer, again, from my last entry: If your position is that AI has no place in the world, a conversation grounded by those parameters would be a very different conversation from the one I’m inviting you to join. If you absolutely hate AI in all applications, you won’t be helping me (or you) to get the nuances right in this tough issue. I respect your opinion but don’t need to read about it and can’t benefit from considering it here.
Whisper House Press 
Leave a comment