As you know, if you’ve been following this podcast for long, my entire brand is around helping you feel good about your writing—not only by developing the actual writing skills, but also by developing the mindset and context skills that go around those writing skills to give you the confidence you need.
And that’s why today I wanted to dig into the issue of uncertainty and how it can pull us toward AI.
We all need a little bit of a pep talk here and there. So today I’m going to give you six mindset reframes that you can use if you’re tempted to use AI or an LLM in your writing.
Why We Feel Tempted to Use AI
The other day, I was scrolling through social media and saw a graduate student I had worked with years and years ago when I was a professor. She had posted something that I think she meant to be cute, but also a flex. She said, “I am good at everything I do.”
Now, I love to see women stepping forward and claiming their expertise. Y’all know this about me. But one of the things that really struck me, partly because of where I am in my own development journey, was that when we are actually good at everything we’re doing, it’s probably time to step a bit outside of our comfort zone, right?
So what I commented back was, “I think it’s time to pick something to be mediocre at for a while, so that you can be really good at that too in the future.”
And this is really the mantra of so many of the folks that I work with at Page & Podium. What I started to think about was how scary, how bold, and how important a move it is for us to seek out a new skill that we feel really, really insecure about.
It is the most human thing in the world to want to stick in your safety zone, your comfort zone, and make sure that you are not pushing yourself out into anything that our reptile brains are going to see as a threat.
One of the things I’ve realized recently about AI is that when we think about how it intersects with our insecurity, what I’m seeing so often is that authors are most tempted to use AI at the point when they feel they’re not yet prepared. They don’t have the skills necessary yet to do the thing that’s being asked of them. We see this a lot when authors are struggling with questions about whether they’re expert enough to write a book.
And in that case, LLMs come forward and say, “Absolutely, I will help you. We can do this. We’ll do it together.”
Y’all, it’s the most natural human thing in this world to want to keep ourselves stuck in our comfort zones. Our literal bodies and brains are programmed to keep us safe by keeping us in familiar spaces, in that zone of proximal development that makes our bodies feel safe but also keeps us from really growing and developing the way we want to.
As we launch into the final episode of our four-part series on AI, I wanted to reach into the human interaction that’s happening with an LLM.
What I’m seeing with a lot of the folks I talk to who have used LLMs in their writing is a real ambivalence. On one hand, they feel uncertain about some aspect of their writing. Whether that’s the sentence level, the structure, or how to organize their ideas, that insecurity is rubbing up against this promise that AI makes to us: I can do it just as well as you can.
This is also why I’ve talked before about how mindset can hold back your writing.
But there’s an ambivalence because, underneath the surface, I also see that authors really, in their gut, don’t want to go the AI-generated route.
Most of the folks I work with would really prefer to push themselves out of their comfort zones and learn this new skill. Many of them are working on their first book, but we know it’s going to be the first of many because oftentimes this is the start of a new career direction, a new branch of your career, a new type of growth.
And, of course, it’s also a big part of soul-searching, of figuring out what your purpose is at this moment.
That’s a really uncomfortable place to be. On one hand, we have AI promising us that we can do these big new things without having to put ourselves out into what feels like dangerous territory. But at the same time, part of the reason we want to write a book is that we want to learn to do these big new things.
I know all of us have moments of weakness. All of us have these moments when we just want to throw our hands up and say, “You know what? If this is gonna get it done, let me just use this. I give up.”
I hate to hear it, y’all.
That’s why today I want to give you six pretty quick mindset shifts that you can use when you feel yourself being pulled toward a shortcut.
These are all specific to AI and LLMs, but you’ll also see that they’re not only specific to that particular shortcut. As we’re working through the process of learning to write our first book—which is, by the way, the first step in learning to write books in general—we’re also learning to get our message out there and to hold the discomfort of the visibility of a book. All of these great things are packed into the process.
And there are so many temptations along the way, whether those are LLMs, AI, or something else.
So keep these in your back pocket. If you’re somebody this is really resonating with, if you’ve felt yourself pulled toward Claude or ChatGPT or whatever it may be, I strongly encourage you to choose one or two of these that really resonate with you. Write them on a sticky note and put them on your computer.
It’s such a scary thing for me to see people talking themselves out of something I know they want to do.
Now, do I have an ulterior motive for this advice? I sure do!
Page & Podium is gearing up right now to welcome our new cohort of published authors. We accept submissions at any point, and we review them through a process called a Manuscript Review.
We’ve structured our Manuscript Reviews specifically to allow you to get insight into whether we’ll be a good fit as your publisher, both from our perspective and yours. But our reviews are also quite thorough, something like a developmental edit that will really help you pull your manuscript forward.
If you have an existing manuscript that’s about 90% of the way done, please reach out. I would love to see what you’ve been working on.
Okay, with that said, let’s dive into the six mindset shifts.
1. You Can’t Just Edit AI Out of Your Manuscript
This first mindset shift is something I’ve found interesting and surprising to follow in industry news: not only can you not just edit AI out of your manuscript, it’s also going to cost you significantly more if you try to do that.
We’ve got two parts here, so let me dig into each of them in turn.
First of all, why am I saying you can’t just edit out AI?
Well, friends, I’m telling you this from experience.
I’ve gotten a fair number of AI-generated manuscripts across my desk. In the spirit of being very open-minded, particularly in the early eras of AI, I read a lot of these and went ahead and took on some of these folks as clients.
One thing I learned very, very quickly is that it’s a much more difficult task as an editor to read AI manuscripts.
Your eyes go quickly over them in a way that means you have to keep rereading sections. That’s not common for me when I’m doing a manuscript review of a human-written text. There seems to be something about the texture, or lack of texture, of the AI voice that really makes it slow.
But the bigger issue is that when we get in there and start deleting the redundancy, deleting the wordiness, deleting the lists of synonyms—all of these things that LLMs pretty much always include, at least to some degree—what we usually find is that there are about three times as many words as you actually need to say something.
In fact, we’ve talked about those lists of synonyms and other patterns in AI-generated writing in our episode on sentence-level editing.
Now, what does that mean for editing?
Well, it means mostly I’m just deleting stuff.
This is not the way we’re going to move forward to a book. What I’ve found across the board is that we need more stuff from the author.
There really is no point at which we can just take an AI-generated manuscript and make it into a human-generated manuscript without basically starting over from scratch.
So the first thing is that it isn’t even really possible to edit an AI manuscript, and I know most folks would not be happy with the result if they tried to do that.
AI-Generated Writing Can Cost More to Edit
The other piece that’s been so interesting is that I’m starting to see a trend, one that I suspect will pick up: if there’s AI-generated content in your manuscript, you’ll be charged an upcharge or surcharge for editing.
I’ve actually seen this up to double the price.
The reason I want to bring this up is that I often find when folks are trying to use these kinds of shortcuts, these ways around our uncertainty, they’re trying to save a little money.
By using AI rather than, say, hiring a writing coach or book coach, you think you’re going to save money because you’ve got this low-cost option.
I really want to make sure everybody hears me say: it’s not actually going to save you money in the long run.
And again, you’re also undermining your ability to really benefit from the book in the way that I know you want to.
So my first mindset reframe, when you’re tempted to write with AI, is to keep in mind that in the editorial world, you’re not going to save money. In fact, what you’re asking for may not even really be possible.
You may be sinking a pretty good amount of time, and depending on your plan, a good amount of money, into an LLM-generated manuscript that then just has to be started over again.
Let’s not do that.
Let’s keep in mind what’s going to happen down the line and recognize that we need to be putting our own human stamp on it anyway.
2. Most Publishers Won’t Take AI-Generated Manuscripts
My second AI reframe is one that you’ve no doubt heard or thought about before.
This is another tip that’s way down the line if you’re just working on your manuscript now, but I would be remiss if I didn’t bring up the point that most publishers will not take your book if it even has the perception of AI.
Now, there are some issues with the way this is being looked at.
We’re starting to be a little more critical in some editorial circles about programs like Pangram, particularly because a lot of the recent scandals involving authors losing book contracts because of Pangram—or, just as often, readers flagging AI content—have affected authors of color.
Most of those high-profile cases, 100%, have affected authors of color.
I think we do need to be very careful about using something like Pangram, where there’s a pretty major stockholder stake in making sure Pangram is very popular.
At Page & Podium, we’ve really stepped back from using Pangram.
We do recommend that authors use it if they’ve found themselves in a place where they can’t really remember where they put AI stuff in. We don’t use it to screen.
But what I’ll tell you is, I actually don’t really need Pangram.
I’ve been reading manuscripts for a really, really long time, and I’m pretty AI-literate. So I’m pretty good at spotting when you’ve used AI.
There have really only been one or two times in several years of folks using AI that I genuinely think I might have been wrong about something identified as AI. In almost all other cases, the author actually admitted to me that they had used it.
It’s not as hard to spot if you’re somebody like me, reading a lot, a lot, a lot of books.
The reason I bring this up is that I often find authors who aren’t saturated in this book world won’t necessarily be able to tell even their own writing from AI-generated writing, at least not across the board.
And this creates a huge problem for you as an author.
What’s going to happen when you’re submitting this manuscript and hearing back that it has AI?
Now you’ve cut your chances with that particular publisher probably forever. Now you have to wade back through and figure out how to get the AI out.
AI-Generated Writing and Copyright
Even if you self-publish, you’re very likely to run into some issues, at least with your copyright.
For a while, it was looking as though this situation where we couldn’t copyright AI books was maybe not going to pan out on the ground.
It was a little bit of a question: Well, how would the Copyright Office know?
But we’ve now been hearing reports from authors who have submitted their manuscripts for copyright and been emailed directly asking whether they used AI.
So this is not something we’re just going to brush to the side. It’s going to continue to be an issue.
In fact, copyrightability is one of the major reasons that publishers don’t want to mess with AI-generated manuscripts. If it’s not copyrightable, it’s not really a piece of intellectual property.
We need to be careful about that as well, not only avoiding using AI in the first place if you’re wanting to have a publisher, but also making sure that if you have used it, you’ve been careful about pulling it out and replacing it with your own writing.
I talked more about how AI affects different stages of publishing in the second episode of this series, AI in Book Publishing.
So my second mindset shift is to keep in mind that if you’re using AI, it’s likely to be very, very difficult for you to find a publisher for that book.
3. Using AI to Write Your Book Stops Your Growth as a Writer
The third mental reframe I want to offer you is something I talked about at length in the first episode of this series, Using AI to Write a Book.
When you’re using an LLM to generate your writing, you’re really short-circuiting your ability to learn.
As I mentioned in that first episode, all writing really is a series of choices. And when we get better at writing, what we’re really getting better at is making choices.
As with anything else that involves a lot of discernment, the only way we get better at that process is by doing it.
There’s unfortunately no way out but through.
When you use an LLM to sidestep the issue of needing to make those choices yourself, you’re not only going to end up with a manuscript that isn’t your choices. You’ve also sidestepped your ability to get better at making those choices.
This, to me, was the reason I stopped using AI for anything related at all to my creative work.
I played in it because I think it’s interesting, and I think it’s part of my job to know what’s happening in that world.
But at the same time, I’m absolutely not willing to sacrifice any drop of my precious creative energy into a robot.
Just not gonna do it.
If it’s going to stop my growth, that is where I draw the line.
I would encourage you to think about that too. Is it really worth it to get this manuscript done more quickly, to feel a little bit more confident about it, at the expense of losing all of the learning you would do by pushing through the discomfort?
I really believe that it’s absolutely worth it to be not that great of a writer for a little bit, so that you free yourself and become independent of these types of hacks.
Then you’re able to write your next book without having so much discomfort, without even having the temptation of using an LLM.
I can tell you, I’m not tempted at all to use an LLM for my writing because I’m a better writer than AI.
But I wasn’t born that way. This isn’t a natural skill.
The reason I’m a better writer than Claude is because I write every single day.
If you sidestep that process of working through the discomfort, you’ll never become a better writer than Claude. You’ll always be inferior to what you’re seeing on the screen coming out of an AI, at least in your mind.
We’ve talked about this before in AI Is Killing Your Writing, and it’s one of the reasons I care so much about authors developing a writing practice that works for them. If that’s something you’re struggling with, we have an episode on finding a writing routine that works.
So my third point is that when you’re short-circuiting that process, you’re stopping your own growth.
Yes, this particular manuscript might get done faster. But to pull in my previous points, you won’t be able to just edit that AI out, which means you’re going to struggle to find a publisher.
You probably won’t even be able to copyright that work as intellectual property.
Bad news bears, y’all.
4. Copying and Pasting AI-Generated Writing Is Plagiarism
My fourth mindset tip on how to think about AI when you’re tempted to use it is something I haven’t heard talked about as often as I think it should be.
I’m just going to say it.
When you copy and paste something out of an LLM, that is quite literally plagiarism.
Now, I understand that I’m going to get pushback here because folks feel like writing the prompts means that they’ve written the actual words.
But I want to go back in time a little bit.
Remember when you were in school and learning what plagiarism was? Weren’t you literally told that it was using words you didn’t yourself write, copying them and pasting them into a document?
That’s what I learned plagiarism was.
To me, even though I get that we have a nameless, faceless author we’re stealing from in Claude or ChatGPT, that’s not actually the case.
Not only were actual authors’ words used to “train” those models, meaning that when you pull words from an LLM, you’re pulling from a database of other people’s written words, but it’s also just not the case that you wrote it.
Prompting someone to do something, whether that’s a robot or a human, is not the same thing as doing it yourself.
And I want to encourage you to think about that a bit.
When you’re copying something from the screen of your LLM and pasting it into your manuscript, you’re doing something that you used to know as plagiarism.
So what’s changed?
We’ve all got to make our own ethical decisions about how we’re going to exist in this AI world.
I know I’ll get pushback on this one, but I want to make sure no one hasn’t thought of that and moved forward, only to realize, “My gosh, what did I do? I ruined this beautiful project by doing something I learned I shouldn’t be doing in seventh grade.”
All right, I know. Tough love in this episode, y’all.
But I’ve got two more tips, and I think they’re going to be really helpful.
5. Using AI Reinforces the Idea That You Need It
When you’re using an LLM, when you’re using any kind of cheat code, I want you to think about how you’re reinforcing the idea in your brain that you need those cheats.
A couple of weeks ago, I used an example someone had shared with me that I found so, so helpful.
What this fellow said was, “Do you remember back before we had GPS in our cars?”
Most of us knew basically our way around at least our neighborhoods, if not our whole towns.
Well, now that most of us have these little handy pocket GPSs, isn’t it true that a lot of people are very, very dependent on GPS to get where they’re trying to go?
Isn’t this exactly the same way it’s working with AI?
When we’re offloading things to AI, we’re basically training ourselves that we don’t need to know how to do those things.
We’re basically letting our brains off the hook.
And when we’re continually giving our brains the message that we’re not able to do this thing we want to do—write this book—we’re effectively reinforcing our inferiority over and over and over again.
Now, if this is something that’s okay with you, if you’re somebody who’s not really interested in adopting the identity of an expert author, if you’re not really interested in stepping into the world of books and the book industry, that might be okay with you.
But most of the folks I work with really do want to set themselves up for future success when it comes to this intellectual life of the mind.
And anytime we’re offloading into an LLM, we’re reinforcing in ourselves that we don’t deserve to do that. That we’re not somebody who can fit in that world, that we shouldn’t even be trying, and that basically we’re inferior to this computer.
Let’s not do it, y’all.
It’s not true.
6. Ask Yourself Why You’re Tempted to Use AI as a Shortcut
My sixth and final reframe is this: when you’re tempted to use AI, keep in mind that using an LLM to do your writing is effectively a life hack.
Now, why do we use life hacks?
Typically, I find we want these shortcuts when it’s something that we, number one, don’t want to do and, number two, don’t care that much about.
And if you’re being pulled toward that LLM, there’s probably something deeper underneath.
There’s probably something that’s making you feel miserable about your writing.
That could be either of the two things I’ve already mentioned: that sense of inferiority, that you don’t belong in this thinking, writing, creating space, and/or the sense that you don’t actually have the skill set you need to do what you’re trying to do.
And it would just be quicker, more comfortable, to get the LLM to do the work for you.
It can be any of those things.
But the thing I want to point out is that until we’ve solved this underlying issue that’s driving you toward the LLM, you’re going to continue to struggle.
Even if it’s not an LLM that you end up leaning on, there will be something that’s going to creep up until we address that underlying mindset, until we really think about why we’re tempted to use a hack.
The people I work with tend to really, really care about their books deeply.
I often have people cry on calls with me because they’re so, so connected to these books.
It’s what I love about my job, honestly.
So that’s not particularly compatible with a life hack approach.
The thing I want you to ask yourself is: Why am I being driven to use a shortcut on something that I want to do and something that I care very much about?
Why wouldn’t I want to do it the right way?
And y’all, I know you do.
Final Thoughts on Using AI to Write Your Book
Okay, this is the wrap-up of our four-part series on AI.
If you’ve missed any of our previous episodes, hop backward in time, either on YouTube, on any of the podcast apps you like to listen to, or at our blog.
As you’re going through, I want you to be really critical. Really think about what AI is doing for you and why it’s pulling you toward it, if in fact it is.
If you’re not a person who’s been tempted by AI, that’s great too.
But please do share these episodes around.
I think we need to be having these conversations about AI in a way that acknowledges the gray area, in a way that acknowledges the messiness of our humanity as it rubs up against the really squeaky-clean tidiness of AI.
I think we’ve got to talk about it.
And we’re never going to talk about it as long as we’re so black and white in our culture about AI good or AI bad.
AI is really complex. There’s a lot to learn, and there’s a lot more that’s going to be coming down the pike.
I think we all need to be investing in this conversation, figuring out where we land, and being thoughtful and generous as we engage with others about the way we feel about AI, the way we’re using it, or the way we’re not using it.
Just as another quick reminder, if you’re somebody who’s 90% of the way through your manuscript or thereabouts, please bookmark pageandpodium.com/manuscript, because I would love to take a look at your manuscript.
At that link, you can find out more about our Manuscript Review. This is the service we offer where we go through your manuscript and give you a very, very clear list of how to fix the major disruptions and deal breakers.
This is something that’s really going to accelerate your progress and make that book so much better.
And, of course, this is an opportunity to pull in a collaborative team where it’s not only your human mind, but all of our human minds thinking together, working together to get your message out there.
If you’re still figuring out where you are in your book journey, you can also take our free Book Strategy Quiz. Answer the same 10 questions that I ask folks on our one-on-one calls, and I’ll tell you via email how to move forward next, along with a custom podcast playlist and free and low-cost resources you can use to unstick your manuscript and move forward to publication.
I hope your manuscript is going well. I can’t wait to see it.
And until then, happy writing!
