How to Make Productive Use of ChatGPT: Q&A with Elisa Lorello

Photo of author Elisa Lorello with the quote: "ChatGPT has fueled my love for writing and being a writer, which is really saying something given how much I already loved both. I see what AI generates and it makes me want to write better, more creatively, and more productively."

Author Elisa Lorello’s exploratory dive into ChatGPT led her to discover its usefulness to fiction and nonfiction writers, and she shares them all in her new book The AI Author Assistant: How to Use Chat GPT to Optimize Your Writing Progress and Income While Retaining Your Human Touch.

In this Q&A, she reveals some of the terrible titles it spat out (while keeping the really good ones to herself), acknowledges the complicated moral question of writers allowing ChatGPT to do their writing for them, and lists the many ways AI has been helpful to her both creatively and practically.

Elisa Lorello (@elisalorello) is the bestselling author of twelve novels and one memoir. The youngest of seven, she grew up on Long Island and graduated with two degrees from the University of Massachusetts-Dartmouth. Since 2010 she’s sold over a half-million units worldwide and has been featured in the Charlotte Observer, Woman’s World magazine, Rachel Ray Every Day magazine, The Montana Quarterly magazine, and Writer’s Digest Online. She’s also been a guest on multiple podcasts and a panelist at BookExpo.

Elisa is a lifelong Duran Duran fan and a proud Gen-Xer, can sing two-part harmony, and devours chocolate chip cookies (not always at the same time). She currently lives in Montana with her husband (bestselling author Craig Lancaster) and their two pets.

KRISTEN TSETSI: Multiple articles about ChatGPT and writing will take one side or the other: they’ll explore writers’ fear of ChatGPT, or they’ll assure writers they shouldn’t fear it at all. As a writer, before ever using it or reading much about it—just knowing that this AI could churn out written material—what were your initial, knee-jerk feelings about it?

ELISA LORELLO: My knee-jerk reaction was the same as many writers: This is not good.

Everywhere I turned, there was an article or a lecture depicting the dangers and negative aspects of AI—people submitting queries to agents that were written by ChatGPT, ChatGPT-generated Buzzfeed articles, etc. I even came across an online course called “Write a Book in 24 Hours with ChatGPT”—I was so curious and skeptical that I purchased it. (It wasn’t a big investment; that said, I do not recommend others buy it. Let’s just say I took one for the team.)

But as a result of that course and experimenting/playing with ChatGPT, I found ways in which ChatGPT could actually be helpful to myself and other writers, and that’s ultimately what inspired me to write The AI Author Assistant. I see it as a little guidebook, a demonstration of as well as a conversation with ChatGPT. I wanted to counter the doom and gloom with an alternate perspective.

In an article in The Independent about the hundreds of AI books that have already appeared on Amazon, Authors Guild executive director Mary Rasenberger says, “This is something we really need to be worried about, these books will flood the market and a lot of authors are going to be out of work.” She then expresses a concern that AI could take the place of existing human ghostwriters, turning “book writing from a craft into a commodity.”

I would argue that book writing is already a commodity more than a craft, in many cases (I ashamedly used to do some commodity co-ghostwriting/editing), but I’m curious to know your thoughts about this: what’s the worst that could happen, do you think?

I think people were sounding the same kinds of alarms about ebooks and the Kindle in 2009–2010. They said ebooks (especially self-published ebooks) were going to kill the printed word and put traditional authors, agents, editors, and bookstores out of business. Digital publishing-on-demand was disruptive, and the industry needed to adjust and adapt.

But here’s the thing: it did. The industry adjusted and adapted, and digital publishing-on-demand is as viable an option as traditional publishing. Moreover, the professional standards for self-publishing significantly increased as a result.

Yes, AI will cause disruption. Yes, there will be a lot of AI-generated books (like, a lot) and content, and yes, there will be a lot of crap to sort through. But I think a savvy reader will know the difference between AI and human, organic writing, and I think publishing will adjust and adapt. Some people will game the system and get rich at the outset, and then the system will catch up and weed them out. Standards will evolve. Because if there’s anything we learned from the rise of ebooks and self-publishing, it’s that fighting or resisting is futile. Downplay the bad and work with the good.

I also think a fraction of the Literary community (capital letter intended) will always rail against genre fiction and mediums like Patreon or Substack or Buzzfeed as “commodity” writing. And hey—I want to make a sustainable living doing this thing that I love. I also want to write the best books and blog posts that I can. Those two things can co-exist.

What led you to see that ChatGPT could be not the nemesis of the writing and publishing community, but an aid to enhance a writer’s existing creativity?

When I was suddenly so creatively fertile! I confess: At first, I was totally looking for “the secret formula” to a bestselling genre novel, and I was curious to see if AI could actually write one. Yet of all the possibilities ChatGPT generated, only one piqued my curiosity. Also, it revealed no secret, just what’s already been said and done. But after a few days of exploration and playing around with titles, tropes, and storylines, every morning in the shower a completely new idea would come to me—and they were good! Not generic, and not necessarily from the AI-generated lists. Moreover, I wanted to experience the pleasure of writing these books myself.

Meanwhile, administrative things I began using it for—outlines and timetables and daily schedules, mainly—were freeing me creatively and improving my productivity and time management. This past month, I started writing two novels with overlapping storylines, kind of like companion novels. ChatGPT generated outlines for each, and I’ve been writing both manuscripts as if they were one novel with alternating POVs. In three weeks, I drafted 35,000 words (combined). At this rate, I predict I’ll complete the first draft of both by the end of June.

Moreover, when I finish writing these two books, I know what the next two will be. It’s like I suddenly gained an edge in productivity, organization, and creativity.

If ChatGPT could do that for a writer like me, then perhaps it wasn’t all bad.

What was your first exploration of ChatGPT like, user-wise? What were your first experiments, and what were your early thoughts about its capabilities or limitations, if you saw any, in terms of how it responded during those first experiments?

At first, I put that course to the test and tried to compile content for a nonfiction book just to see if it actually could be done in the alleged 24 hours. It generated about 10,000–15,000 words of content—hardly enough for the kind of book I had in mind for the test case. Moreover, I noticed factual inaccuracies in some of that content. (ChatGPT sometimes doubles down when you challenge the inaccuracy; it’s a little creepy.) And most AI-generated writing isn’t very good from a craft standpoint.

Next, I did some creative exploration for fiction—titles, story ideas, and even scenes of description and dialogue. For example, I asked Chat GPT “What kind of character would appeal to a Generation X female reader?” Or “What are the most popular tropes for contemporary romance?” When I wrote The AI Author Assistant, I asked for title and subtitle recommendations. In the book, I showed the progress of how I ultimately came up with the title I did.

The most interesting and unexpected result was all that exploration and play sparked ideas of my own. So, for example, if I asked ChatGPT to give me 10 premises for an office romance, I would decide they were all too generic—and then the following morning in the shower a fresh idea for an office romance would come to me.

I also saw ChatGPT’s value in generating story outlines. This was one of the game-changers for me. Writing outlines before I start a novel has always sapped me creatively. By giving ChatGPT the premise at the onset, along with characters and any scenes I might already have in mind, ChatGPT then produced an outline within seconds. Now I had something to work with at the beginning of a manuscript, and I was creatively motivated rather than sapped. And even though the outline tends to be as generic as much of the content ChatGPT produces, it’s fluid and flexible, and place-holds plot points for me. I’m OK with that.

Ultimately, ChatGPT can function as a virtual assistant—you can use it to brainstorm ideas, write outlines, and even prioritize your day’s to-do list. It can also help you with basic research and will deliver it to you quicker than a Google search.

However, beware—as I mentioned earlier, the information isn’t always accurate. In The AI Author Assistant, I included an example of my asking ChatGPT to rank my books from most sales to least sales. It listed seven books, I think (I’ve written over a dozen). The first one was my least-selling, and of all the titles, only three were actually mine.

Additionally, if you ask it to actually write a scene or paragraph for you, be prepared to do a lot of editing and polishing because the prose is wordy, the dialogue is stilted, and the description is cliché and generic.

To offer my own example of ChatGPT doubling down on errors, my husband Ian asked ChatGPT, “Who is Kristen Tsetsi?” After listing some basic information available online, it said I was also the founder of National Freelancers Day. Ian then asked, so it might recognize its error, “Who is the founder of National Freelancers Day?” ChatGPT said, “After double-checking, I can confirm that Kristen Tsetsi is indeed the founder of National Freelancers Day.”

Back to your experiences with it. You write that ChatGPT can “streamline your writing process and progress by generating text quickly, reducing the time spent typing or brainstorming.”

What does that mean? I ask because it reads as if the AI is in fact doing a portion of the writing, and my reflex as a writer is to think that any writing not done by the writer is a form of cheating. (Or plagiarism? Is it plagiarism if the writer is a bot?)

I think ChatGPT wrote that line, actually. [winks] I asked ChatGPT to give me ideas for the ways in which writers can use ChatGPT ethically and responsibly and then applied the responses to The AI Author Assistant as a kind of dialogue.

I think your interpretation isn’t wrong—it could be used that way. And I share your concerns about plagiarism and/or cheating. I wrestled with those questions in the book. Is ChatGPT serving as a ghostwriter? If you cut and paste its content and pass it off as your own, is it plagiarism? Are you cheating your readers and yourself out of an authentic writing experience? I didn’t provide any specific answers because I don’t think they’re hardcore “yes” or “no.” My main goal was to bring the questions, concerns, and discourse front and center.

I use ChatGPT as a springboard. For example, I dislike writing book descriptions, and I always freeze up when it’s time to write one. I asked ChatGPT to write a book description for The AI Author Assistant. I hated what it came up with; however, it unblocked me and I wrote a description on my own. (I used only one line from the AI-generated one, and tweaked it a bit.) I did the same writing copy for Amazon ads. ChatGPT gave me some ideas to work with, and I then created copy in my own words.

In other words, ChatGPT is a buddy. For instance, instead of asking my husband (also a writer, and a busy one at that), “Can you help me brainstorm book titles?” or “Can you help me write this description?” or “Can you help me flesh out this scene?” ChatGPT can serve in that role.

And when it comes to prioritizing a to-do list, ChatGPT saves me a lot of time. I type in the tasks and it not only prioritizes them but also explains why it chose the order it did. Ditto for the outlines. ChatGPT can generate an outline in seconds; it takes me at least a couple of hours (and that’s just the first go-round), and after I do it, I never want to go near the story again.

What command did you use to ask it to write the book description? If I wanted ChatGPT to write some back-cover text (even if I intended to heavily edit it), I would have no idea what to tell it to make it do that. How much should be said about the book? Can you provide command samples for fiction and nonfiction?

For The AI Author Assistant, I gave ChatGPT the premise and objective of the book, as well as keywords that would likely come up in an Amazon search. So I think the command was something like: “Please [I’m very polite] write an Amazon product page description for The AI Author Assistant that highlights the objective of writers using ChatGPT in ethical and responsible ways, and include the following keywords…”

For fiction, I would probably do something similar. “[Insert title] is a contemporary romance in which four friends who haven’t seen each other in 20 years reunite at their hometown mall set for demolition in one week. Please write a description for the back of the book and the Amazon product page that uses these details and keywords…”

I would also give it a word count as a guideline.

ChatGPT tends to write kitschy copy like “Do you love a contemporary romance about second chances? [Title] has quirky characters, second chances, and hearty humor!” You may have to tell it to tamp down the selling while also appealing to the reader.

If someone is writing a novel—let’s say it’s literary—and they’re stuck on the “what next?” of a scene, can ChatGPT help?

I think it can if you use it as a freewriting technique. For example, if I don’t know what scene comes next, I could summarize (or perhaps even copy and paste) the previous scene and outright ask ChatGPT “What do you think should happen next?”

In the past, I’ve tried to unblock myself by typing, “What I’m trying to say is…” and then proceeding to try to work it out on the page, however messy it may be. You can say that to ChatGPT and it could potentially help you organize your thoughts or give you clarity or direction.

The key, I think, is to always make it your own. I’ve gleaned ideas from my husband or my former co-author or a developmental editor, for example. I’ve written dialogue after overhearing two people in a coffee shop or using a conversation I had verbatim. Writers also rely on beta readers for feedback. We’re pulling ideas and information from so many different contexts. It’s how we re-contextualize them that matters. We always need to be responsible and ethical.

In your book’s intro—and the part I’m referring to was written by ChatGPT, which is clear because of the different font you assigned it—there’s a claim that ChatGPT can “help you create marketing and promotional campaigns and materials which, in turn, can increase book sales.”

What can ChatGPT do for authors who are already doing all of the regular things they’ve been told to do to market their work?

I haven’t fully tested its usefulness, but ChatGPT purports to help you generate content for social media marketing, mostly. Also write press releases, book blurbs, ad copy, blog posts, etc.

I’ve been mostly experimenting, asking it to give me sample tweets to promote my most recent novel or brainstorming content ideas for Instagram and/or Facebook. I haven’t applied any of it yet (other than the aforementioned Amazon ad copy). I’ve also outright asked ChatGPT things like “What are some unique ideas for authors to market their books?” or “How can I make my monthly newsletter content engaging besides the usual suggestions of sharing excerpts of my WIP or photos of my workspace?”

Other than ideas for swag, I didn’t see any suggestions that made me go, “Now that is something different.”

I asked for ideas for a Substack series the other day; I might consider a couple of those.

Honestly, most of what ChatGPT suggests has been said and done before. But something about it sparks my creativity and originality. I reject a lot of what ChatGPT suggests and find better ways to write or think about an idea or solve a problem. I like that kind of mental exercise. And if it’s suggesting things I’m already doing, then maybe I’m on the right track and I don’t need to change what I’m doing but rather how I’m doing it.

Many writers get stumped on book titles and would probably love any kind of help they can get. But were any of the title offerings ChatGPT gave you absolutely terrible?

Oh, definitely a lot of terrible titles! But it’s like sifting through multiple pans of dirt for one gold nugget. The fiction titles are especially bad, although I confess there was one that I loved and am going to use in the future (I’m not telling you what it is!).

The advantage is that these lists of titles can be generated in seconds. That’s the “streamlining” of progress and process. I don’t need to type or brainstorm the lists myself. But if I sift through two or three sets of AI-generated lists, I’m bound to spot a nugget I can work with.

For example, the other day I got an idea for a “Freaky Friday” time-travel novella. ChatGPT generated a decent outline, but here are the titles it suggested:

  1. “Mom Swap”
  2. “Back to the Past”
  3. “Freaky Family”
  4. “The Mother-Daughter Time Warp”
  5. “A Switch in Time”
  6. “My Mother’s Shoes”
  7. “Reversing Roles”
  8. “The Time-Travel Experiment”
  9. “Two Generations Apart”
  10. “A Tale of Two Ages”

I mean, those are cringe-worthy. I would either request an additional ten after I’ve developed the story or take another long shower and wait for a better title to come to me.

What key cautions would you give writers about using ChatGPT?

I’ve mentioned a couple. For one, the information it feeds you isn’t always accurate. In fact, ChatGPT sent me on a wild goose chase to press releases and webpages that downright didn’t exist. You can’t take the information it provides you at face value.

But if you want to know something like the differences between Millennials and Gen-Z or the characteristics of a particular genre, then it’s a good, quick resource. The other day, I asked ChatGPT how one might go about opening a particular kind of business in a particular location for a potential plot point. It gave me enough background that if the character was going to pursue it, then I could write it in a way that was plausible.

The other caveat is the plagiarism issue we discussed earlier. Where’s the line, and how blurry is it? Is it OK for me to post ChatGPT-generated tweets, for example? Would anyone notice or care? (Maybe I should try it as an experiment!) Is it OK for me to use AI-generated ideas and titles for a blog series, but write my own content? Is it OK for ChatGPT to ghostwrite the content for a site like Buzzfeed, for example, while I collect the paycheck and the credit?

I posit that while we navigate through this technology and its applications, at the end of the day you need to be your own moral compass. And if your compass differs significantly from the industry’s, then you need to take the consequences if they reject it.

Cover of The AI Author Assistant by Elisa Lorello

Has using AI taught you, or encouraged you to see in a new way, anything about your own writing?

It’s fueled my love for writing and being a writer, which is really saying something given how much I already loved both. I see what AI generates and it makes me want to write better, more creatively, and more productively. It makes me want to be more professional and organized. Ultimately, that’s the feeling I wanted readers to leave The AI Author Assistant with.

Overall, I encourage writers to play with ChatGPT and approach it as such. Have fun with it. Be curious. Once you know what it can and can’t do for you, I think you’ll have a better sense of how to use it in practical and responsible ways.

Share on:
Notify of

This site uses Akismet to reduce spam. Learn how your comment data is processed.

newest most voted
Inline Feedbacks
View all comments