Using AI Without Compromising the Story

By James M. Jackson

I belong to several writing groups that have online forums for their members. Recently, a highly respected member on one of the forums went on a rant about how any use of large-language-model artificial intelligence (like Claude, ChatGPT, and others that analyze text and identify patterns—LLMs, for short) is totally unacceptable for authors. The rationale pointed to LLMs' unethical use of copyrighted material to learn, and that it’s dumb to do anything to help a product that undercuts your product. I’ve paraphrased a much longer and more erudite analysis. In this view, writers should feed the backlash and shut AI down.

Standing on its grandest pedestal, this is the equivalent of saying that eating anything produced by American farmers on United States soil is unacceptable and should be banned because all the land used to produce the food was stolen or otherwise obtained illegally or immorally from the indigenous peoples. What’s more, if we only stopped using this product, farmers would stop producing it.

Of course, we need food to eat and don’t need generative AI to live—at least not yet.

Maybe a better analogy is that writers should go back to scribbling longhand because science indicates that doing so uses a larger percentage of our enfeebled brains than keyboarding does. But I choose to keyboard instead of writing longhand because it’s much faster (and I can actually read the result two hours later, instead of scratching my head about the intended meaning of those ink marks on paper).

My firm line with AI is that I will not have anyone or anything do my creative writing for me. I do accept human and AI suggestions to improve my work, and I will absolutely allow LLMs to produce a first draft synopsis for my 95,000-word novel. And I’ll certainly allow it to generate multiple suggestions for the book’s blurb. Even in those instances where LLMs did the initial draft, the final words and styling will be my choices, even if composed mostly of words chosen by the AI. If AI can help me perform a routine task faster, I’m all in.

Here are three examples of how I recently used Claude to perform structural analysis.

In my current WIP, Niki Unbound (Niki Undercover #3), I realized Cindy (Paddy’s wife and Seamus’s daughter-in-law) went missing for part of the story. In the before-days, as part of the read-through of the entire novel, I would have set up a little worksheet to keep track of where Cindy was and what she was doing. Since I knew it was a problem, I gave Claude this prompt:

Where's Cindy: One of the areas to check is Cindy's location in each scene. Please review Niki Unbound draft 2 2026-02-21.docx and provide a chart with the following columns: 1. Scene number 2. Cindy present Y/N 3. Cindy's location [Chicago area, D.C. area, Shank Lake vicinity, Other (specify), Unknown]. Please provide results in a downloadable Excel file. Questions?

On its own, Claude color-coded the lines to mark location (making it visually very easy to see where Cindy was), provided a short note for each scene, and provided commentary on a few scenes it thought I should look at because they were somewhat ambiguous. The graphic below shows the first thirteen scenes.

With this information, I can pinpoint where to fix this issue and do that before doing my read-through.

The second area I’ve experimented with is having Claude do a first pass at analyzing scenes for their impact on the story. Normally, before I hand the second draft over to Jan, I do a read-through and capture on a spreadsheet a number of characteristics for each scene. I want to know if it’s Action (A) or Reaction (R) or a combination. I also want to know which scenes have increased Twists (T), Dangers (D), and Reversals (R)—a concept I picked up from Jane Cleland. I also record Jess Lourey’s approach: A.R.I.S.E. (Action, Relationship, Information, Suspense, Emotion). After providing the basic definitions to Claude, I asked it to do a first pass at analyzing each scene.

Here’s what it did for the first few scenes.

I’m not claiming Claude got this 100% correct, but this makes it easy to spot potential issues that I can then investigate. It took me only a few minutes to formulate the query and give Claude the manuscript, and it has already saved me hours of work.

The third task was also in preparation for the read-through. Because I write the first draft mostly by the seat of my pants, following characters where they take me, I create plot holes and cul-de-sacs that I must fix in the second draft. Normally, I discover the problems as part of my read-through. This time I queried Claude:

Please review the Niki Unbound draft document for any and all continuity issues. Be thorough. Questions before proceeding?

It produced a four-page Word document with “issues.” It grouped them into Critical – Plot Holes & Logic Gaps, Significant – Technical Errors & Story Problems, Moderate – Tone, Pacing & Character, Missing / Unverified, and (unrequested) Typos & Copy Errors. Here’s the top part of the report:

Claude was good, but not great. It identified sixteen issues, of which four were misinterpretations on its part. An example is #2 above. It correctly identified eight people but missed the ninth. When I pointed out who I thought the ninth was, it checked and agreed I was correct. Back when I was in school, a 75% on a test was a C-minus. Until I do the read-through, I won’t know if it should have flagged other issues.

But even with the mistakes, it found a dozen issues I could fix before I did my read-through. That, I believe—I haven’t done it yet—will make my read-through faster and more valuable.

I’m not trying to convince writers how they should or should not use AI. But if you are AI-curious and tolerant, these three examples may spark some ideas for you. Writers and readers, where is your firm line in using AI?

* * * * *

James M. Jackson writes justice-driven thrillers with brains and bite, including the Niki Undercover Thriller series and the Seamus McCree series. To learn more information about Jim and his books, check out his website, https://jamesmjackson.com. You can sign up for his newsletter (and get to read Low Tide at Tybee, a novella featuring Seamus, his darts-throwing mother, and six-year-old granddaughter, Megan).

Thoughts on Artificial Intelligence by Nancy L. Eady

Last month, I posted a short short story I called “Customer Complaint” on this blog. A few days later. I was searching for something unrelated when a critique of the story I had posted popped up spontaneously—generated by the search engine’s AI program. I’m not sure why. The AI program told me it liked the story, even calling one or two elements of the story “comedy gold.” I don’t know what measure it used to evaluate the story, and because it is AI, and on my computer, I wondered if it gave me its honest opinion (if AI can have an opinion), or if it was just telling me what it thought I would like to hear. 

I don’t trust AI yet. Part of that is my native skepticism of all things mechanical—any office worker knows that copiers, for example, cannot be trusted. Never let the photocopier know you are working against a deadline. Doing so increases the chance of a serious malfunction bringing your work to a screeching halt past 90%. Part of my mistrust stems from my understanding a bit about how it works, while another part stems from not knowing enough about how it works. And as a writer, I have major concerns about how you can control plagiarism of your work by AI. 

The AI that has burst most spectacularly onto the world stage are the large language models (LLMs) like ChatGPT, Gemini, and Copilot. But LLMs are more like probability engines than basic search engines. LLMs look at your question, sort through the many ways other people have answered similar questions in the past, and put together the response that is most likely to be relevant. That doesn’t mean that an LLM will come up with the right answer. The legal profession, for example, is inundated with stories of lawyers who used something like ChatGPT to write a brief and ended up turning in a brief that looked good but had citations to non-existent cases. The term “hallucinated cases” describes this illusory case law. To put it mildly, judges are not amused when lawyers submit such briefs. The sanctions for doing so are becoming increasingly severe. One lawyer I know uses a Chat GPT type platform as an aid, not as a substitute for doing his own research. He tells an amusing story about arguing with the AI program about its hallucinating cases. He told the program it was hallucinating cases, and the program “shouted” back at him (all caps) “I AM NOT HALLUCINATING CASES.” (But it was.) 

The search for the most relevant answer also has another side effect. The one answer an AI LLM program is not likely to give is “I don’t know.” Their programming insists that they provide an answer, and they will do exactly that, regardless of whether the answer provided is correct.  

The part of the technology I don’t understand is what allows AI to have conversations with people. We recently upgraded our Amazon Alexa devices to Alexa Plus, which is AI-based. (We did that to stop fighting with the bedroom Alexa over the differences between the phrase “turn on the bedroom lights” and “turn on bedroom lights.”) I have had two tentative conversations with Alexa Plus where I was asking her/it questions with no right or wrong answers; questions about what she likes or feels, just to see what would happen. (My daughter gave me some weird looks when she walked in on those). I felt as if I were talking to a person.  

I use ProWritingAid to help me edit my work. It also has an AI option, but for now, I leave that feature alone. But the availability of AI writing aids is growing because AI can be useful. That usefulness will only increase with time as the technology develops further. If you can trust the AI program you use to keep your work confidential rather than adding your work into its repertoire of training material, it could be useful in proofreading. It also might be helpful in generating ideas on how to promote your work and in providing research sources you may not have considered.

Are there any ways you use AI to help you with your writing efforts, whether it is in proofreading, editing, promotion, marketing, or something else? If you have tried AI, what do you think of it? What is your general impression of the technology?

Amazon vs Google – AI audiobook narrators

By James M. Jackson

I released Niki Undercover, the first of a new series, using Amazon-only distribution even though my Seamus McCree series has wide distribution. I’d prefer to distribute wide, but to do that well, I would need to spend more time than I had available to market to all channels. Initially distributing only through Amazon allows me to focus my marketing time and, by committing for at least ninety days to stay solely within the Amazon distribution system by joining KDP Select, it compensates me for Kindle Unlimited readers who read the book as part of their subscription without additional cost.

None of my books sell well enough to justify my spending money to pay a human narrator to create audiobooks. Maybe that will be different for the Niki Undercover Thriller series. For now, my choice is to create an AI-narrated audiobook or none. In the past, I used GooglePlay to create audiobooks. KDP (Amazon’s publishing arm for self-publishing) recently expanded availability of its beta version of AI-narrated books. Niki Undercover met their requirements, so I gave it a go.

Having completed that project (the audiobook is available on Amazon), I want to share my thoughts on how Amazon’s AI product and Google’s compare.

Structure and Basic Philosophy

Both providers are in beta, meaning they can and will change their offerings. Google is available worldwide; Amazon only in select countries and for select books.

Both providers start with an eBook and automatically convert it into an audiobook.

Amazon wants the audiobook to comport exactly with the eBook to facilitate those who purchase both and want to switch from one format to another. It converts only the sections it considers the story (ignoring all the other parts of the provided eBook). It does not allow you to add, subtract, or modify text—doing so makes it difficult to sync the eBook and audiobook.

Google converts the entire eBook, including title page, copyright, author’s note, and other works. Google allows you to add or subtract material to create your audiobook as you choose. It also allows you to modify the words of the eBook if you think that improves the listening experience. For example, without using dialog tags, paragraphing alone may allow an eBook reader to know which character speaks each line. The audiobook loses those visual clues. Google allows you to add tags to make dialog attribution clear. Amazon does not allow any changes to the text.

The ability to modify the text gives Google a huge advantage, but one Amazon can only match if it gives up the notion of 100% syncing between the two versions of the book.

Voices

Amazon provides 80 narrator voices covering masculine and feminine traits, age groups 30-40 and 40-50, and American, British, Australian, and Southern U.S. accents.

Google has 50 narrator voices. For English, they have 13 American, 3 Australian, 4 British, and 5 Indian voices. The 25 foreign language voices cover Spanish (Spain and Mexico), French, German, Portuguese, and Hindi. They also provide masculine and feminine voices, but have three age ranges: 18-30, 31-45, 46-60.

Advantage to Amazon unless you don’t want English.

Changing Voices

Amazon allows you to select a different voice for each chapter. Google allows you to change the narrator for the entire chapter, a scene, a paragraph, or even an individual word.

Advantage Google.

Pronunciation

Say this in your head: After I read a piece of fiction, I will have read that piece of fiction. When you read that convoluted sentence, you pronounced the first read as “reed” and the second read as “red.” AI narrators are not that sophisticated . . . yet. And with place names, you never know what you might get. Both Amazon and Google have a mechanism to tell the AI narrator how to pronounce a word.

Amazon lets you choose a word and spell it differently in order to get the narrator to pronounce it correctly. So, if Amazon pronounced both reads as “reed,” I can change the second one by changing its pronounced spelling to “red.” That’s easy, but when I had an organization known by its initials AHI, its attempt was something like “ah-hee.” I’d pronounce each of the letters: A H I. When I tried that, it came out “ah-hi.” My final solution was “eh H eye.” I ran across a couple of words I could not figure out how to trick the narrator into pronouncing them correctly. For those, I got close enough and abandoned the fight.

Once you have corrected a pronunciation, it asks if you always want to use the new version. That’s great for people or place names, but not good for read vs. read.

This ability to change pronunciation solved another problem. I used military time in my chapter subheadings. It pronounced something like Day 6, 2130 EDT as Day six, two thousand, one hundred thirty Ee Dee Tee. I could change 2130 to “twenty-one thirty hours” and EDT to “Eastern Daylight Time.”

Google has a similar ability to change word pronunciation. Rather than using just letters, they introduce their own version of phonetic spelling. Usually, I could type in another word that incorporated the phonetic syllable I wanted to reproduce and splice something that worked. Then, in an improvement, Google added the ability to speak the word as you want it said. It usually reproduced my vocalizations correctly. Sometimes my accent and the AI narrator’s did NOT work, and I was left to figure out the phonetic equivalent or try enunciating more clearly.

Google allows you to apply the new pronunciation universally.

Advantage: Mixed. My impression was Amazon’s AI had fewer words I needed to fix. And Amazon’s English letters were easier to figure out than Google’s phonetic approach. That said, the overall advantage still goes to Google because it allows the voice-recording approach.

Pauses and Dramatic Interpretation

Both AI narrators attempt to interpret the text and read punctuation clues to add pauses. Periods work well, commas are a mixed bag, and question marks and exclamation points can lead to wonky and inconsistent inflections. You can’t fix those, but you can do something about pauses.

Amazon allows you to shorten or lengthen pauses by 25% or 50%. Google does not allow you to shorten pauses (I didn’t find any instances where I wanted to with Amazon). For additional pauses, it offers choices in tenths of seconds from .5 seconds to 2.0.

My sense was Amazon’s narrator interpreted text better than Google. However, I have not listened to Google’s narration in the last year, and it may have improved.

Advantage: Amazon—maybe.

Cover Art

Amazon takes your eBook cover and converts it to a rounded square cover for Audible, by placing the eBook cover in the square and filling in the sides with a contrasting color. No changes allowed. (See left for Niki Undercover example.) If you look for the audiobook through the standard Amazon store, you get the same cover as the eBook. Google asks you to upload a square cover, which you can design.

Advantage: Draw – Amazon makes it easy, Google gives choice.

Where else you can publish the audiobook

Google allows you to take the files they create and publish them anywhere.

Although I’ve read people saying you can publish the Amazon AI-narrated audiobook on places that distribute to libraries, I cannot find that in the Terms of Service. On a help topic it says, “Customers can find and listen to audiobooks with virtual voice wherever Audible audiobooks are available today. This includes the opportunity to reach millions of listeners on Amazon, Audible, Alexa, and through Amazon Music Unlimited.”

Advantage: Google

Pricing and Royalties

In typical Amazon fashion, they require you to price your AI-narrated audiobook between $3.99 and $14.99. They pay 40% royalties. Google allows you complete pricing freedom on GooglePlay, provided you do not offer it at a lower cost elsewhere. They pay 52% royalties.

Advantage: Google

Overall Assessment

Google provides significantly more flexibility than Amazon except for the number of available English-speaking voices. I would choose Google unless (as is my current situation) I wanted to stay within the Amazon environment.

I look forward to your questions and comments.

* * * * *

James M. Jackson writes justice-driven thrillers with brains and bite, including the Niki Undercover Thriller series and the Seamus McCree series. To learn more information about Jim and his books, check out his website, https://jamesmjackson.com. You can sign up for his newsletter (and get to read two free short stories, one featuring Niki and Seamus and the other taking place at Seamus’s camp in Michigan’s Upper Peninsula).

Major Changes in the Last Year

By James M. Jackson 

Three years ago, I agreed to create and teach a course on Indie Publishing for the Guppy Chapter of Sisters in Crime. This Sunday, I’ll welcome my third group of students. In order to give students time to do the reading and suggested homework, we cover the material over six weeks.

I designed the course to provide students with hands-on experience with all aspects of Indie Publishing other than writing/editing and marketing. We discuss the pros and cons of Indie Publishing compared to more traditional publishing paths, business formation, business financials, sales and income taxes, the process of physically creating, perfecting, and publishing print, eBooks, and audiobooks, and much more.

To prepare for this new session, I reviewed the twenty lessons. Some required only minor changes from last year. Changes in four areas, however, required significant reworking of the lessons I used for last year’s class. Three relate to so-called Artificial Intelligence, AI for short. The fourth is the increasing opportunity for selling direct and using bespoke books.

The more I thought about it, the more convinced I became that I needed to create a lesson that provides information on how Indie and Hybrid authors are using AI. With a topic so large, the lesson can only skate the surface, but my sense is that writers who consider becoming Indie or Hybrid authors cannot make a reasoned decision without understanding how AI might affect that choice.

Besides the broad AI overview, my lessons on cover design and audiobook creation required extensive revisions to reflect the significant improvements in AI over the last year. I also needed to reflect information about the increased acceptance of AI technology-assisted narration and cover creation and processes an Indie author might use.

The topic of AI use by writers is fraught with strong emotion. But ignoring it in a class for writers learning the inside skinny on Indie Publishing seems to me a disservice.

Your thoughts?

* * *

James M. Jackson authors the Seamus McCree series. Full of mystery and suspense, these thrillers explore financial crimes, family relationships, and what happens when they mix. To learn more information about Jim and his books, check out his website, https://jamesmjackson.com. You can sign up for his newsletter (and get to read a free Seamus McCree short story).

AI for Developmental Editing

By James M. Jackson

We’ve been dealing with “Artificial Intelligence” for years. The term was created in 1956 at the Dartmouth Conference.[i] Ten years later, ELIZA, created by Joseph Weizenbaum at MIT, used an early natural language processing program (aka chatbot) to simulate a Rogerian psychotherapist![ii] In the 1970s and 1980s, we all started pressing our phones to indicate 1 for sales, 2 for service, etc. I started using a spellchecker in the mid 1980s.

The good old days.

Now, with massive and inexpensive computing power, AI is engaging in a much wider set of tasks. The software program ProWritingAid, which I have used for years, uses AI to check spelling and grammar. It also checks for a large variety of writing “issues” including writing style, sentence length and variety, readability grade, use of passive voice, adverb use, emotional tells, acronym use and consistency, paragraph complexity, and much more.

It not only spots “errors,” it suggests how to correct the “problem.” I use quotation marks because it is not always right. Often its software becomes confused by complexity, and misidentifies things like dialogue tags, doesn’t realize that a word like goin’ uses an apostrophe to signify the missing g rather than claiming my writing is flawed because I ended a quotation without starting it.

Although not perfect, it is very useful. So, when the ProWritingAid team introduced a Manuscript Analysis feature, I was interested to see how it would do. I recently finished Draft 3 of Niki Unleashed (Niki Undercover #2) and knew it still had a few structural problems. Before I tackled those, I asked the Manuscript Analysis to do its thing.

Its report was comprehensive, covering 37 pages. It began by telling me Niki Unleashed is a Thriller (Political Thriller) and provided a story overview and plot outline. The story overview (see below) is accurate, if not inspired. The plot outline caught the chief points but missed the subtlety.

With that start, I was skeptical about how helpful this would be.

But I changed my mind after reading the narrative elements section. It accurately reported three themes and motifs. Fortunately, there was more of this kind of analysis later in the report.

Then it stumbled with its analysis of potential readers that I did not find enlightening. Its “Similar Books” section explained the rationale for its choices, although I did not find the three choices compelling.

It listed characters, their role, character type, story presence, and offered for the top three a similar fictional character. The comparisons were not ones I would have made, but the explanations made sense. A table of contents section that listed narrative person and POV had no value to me.

The real value came in the next two sections.

It uncovered five Narrative Themes and concluded I had used two effectively. The other three themes were underdeveloped. It provided text references for its comments, and for the three “underdeveloped” themes provided suggestions to strengthen them. Two of the underdeveloped themes are underdeveloped because I do not consider them to be major themes. The third, Betrayal and Trust, does play a significant role in the story. Here is its suggestion:

Broad brush, but an accurate assessment and something I can improve in draft 4.

The Plot & Structure section is where ProWritingAid’s AI shined. It discussed seventeen areas in significant detail. Maybe it was trying to sugar-coat the pill, but it reported several things it thought I did well (a compelling premise, clever use of red herrings and suspense in the murder investigation, and fast-paced and action-packed narrative in the second half).

The greatest value was in the things it considered issues.

Several involved inconsistencies it found. While not all of them were actual inconsistencies, the fact it considered them problems means they are things I must clean up.

It was concerned about three plot points it didn’t think I had resolved or had not provided sufficient motivation to justify. (It was right).

Recall that it praised my pacing in the second half of the novel. Issues it highlighted in the first half were an overuse of internal monologue and lengthy explanations of procedures and technical details. Those issues slow the pace.

It also worried about “unclear and inconsistent states and threat levels” and “inconsistent tone and use of humor.” These are two facets of the same issue. One subplot has a different level of threat (personal relations and financial well-being) than the main plot (her life is on the line). While I don’t agree with its suggested solutions, it spotlighted a potential problem I will ask beta readers to express their opinions on.

On characters, it gave me positive marks for a complex and multifaceted protagonist, strong voice and authentic dialogue by the protagonist, and a realistic portrayal of the protagonist’s internal struggles and emotional journey.

It highlighted unclear motivations for two of the other characters—fair assessments, I think.

It also dinged me for “underdeveloped relationships and missing backstory of supporting characters.” This is book two of the series focused on the protagonist. The AI tool hasn’t read book 1, and it wanted more character development for supporting characters. Maybe I should include a few more snippets of backstory, but this isn’t a novel where everyone shares the spotlight and needs to change.

Setting: It praised the use of setting to create tension, but it wanted (depending on the criticism) both more and less detail. It faulted some of the legal procedures I used, as well as some of the FBI/ATF procedures. This isn’t a procedural, so I need to cut corners, but it is an interesting observation that I will think about.

This first use came free of additional cost with my ProWritingAid license. Each future use costs $50. It doesn’t replace a human editor, but I think I received $50 worth of value. My next draft will be stronger based on its findings.

Next month, I’ll talk about how I use ChatGPT to augment this ProWritingAid Manuscript Analysis.

Authors: what are your thoughts on using AI to help improve your stories?

Readers: Do you enjoy learning the behind-the-screen aspects of how we create our novels?

---

[i] https://home.dartmouth.edu/about/artificial-intelligence-ai-coined-dartmouth

[ii] https://www.csail.mit.edu/news/eliza-wins-peabody-award

James M. Jackson authors the Seamus McCree series. Full of mystery and suspense, these thrillers explore financial crimes, family relationships, and what happens when they mix. To learn more information about Jim and his books, check out his website, https://jamesmjackson.com. You can sign up for his newsletter (and get to read a free Seamus McCree short story).

 

Using AI to Help Create a Book Cover

By James M. Jackson

My journey of using personal photographs in my book covers started with a traditionally published book, Cabin Fever. The publisher’s cover artist used a stock photo of a cabin that was much too ornate for Seamus McCree’s guest cabin and worse, since the story takes place in the dead of winter in Michigan’s Upper Peninsula, he included snow . . .on deciduous trees that still had all their leaves. The coup de grâce was strings of lights hanging from the cabin (which has no electricity), you know, the ones that are supposed to make you think of icicles. After a good laugh, I sent him a picture of the actual cabin I had used as a model—which he incorporated into the final cover.

Once I became an Indie Publisher, I decided I would continue to use my photos in the cover creations. Empty Promises was my first try. In the story, Seamus’s granddog finds bones in a derelict cabin. I chose to depict that cabin on the cover. The actual cabin I photographed was in too good shape, so I Photoshopped the image to distress the cabin, sloping the roof and shifting the extension to look as though it were about to fall down. I added a couple of holes in the roof for good measure.

Next came my first novella, “Low Tide at Tybee.” I chose a picture of the Tybee Island beach populated by sanderlings, energetic small sandpipers described in the story.

The action in False Bottom returned to Boston, Seamus’s hometown and included references to cemeteries and a Celtic Cross. I incorporated a picture I had taken at the Gettysburg Battlefield of the monument commemorating New York’s “Irish Brigade.” The actual statue has a life-sized Irish Wolfhound lying on its base that I had to carefully remove. I merged fog from another photo to create the cover.

The novella “Furthermore” story follows immediately after False Bottom. It made sense to me to link the two covers. I used the same elements, but pushed the monument further to the back and more in the fog, which fits the story.

For both Granite Oath and the compendium Seamus McCree U.P. North, I picked photographs of Shank Lake, Michigan, where Seamus has his “camp.” Interesting fact on the Granite Oath cover: given Seamus’s camp faces west over Shank Lake, most people think the cover depicts a sunset; the actual picture was facing west, but of a sunrise reflected in the clouds and lake.

For the last several publications, I have enlisted the help of my Readers Group (newsletter subscribers) and Facebook followers to help select the cover. I present four or five alternative covers, give them a brief teaser about the story, and ask them to choose their favorite and tell me what they liked or didn’t like about each one. Hijacked Legacy takes place at Seamus’s camp in late fall. Light snow falls during a critical scene. Each of the five offered choices had strong proponents and equally strong detractors. All were terrific photos. The problem, I decided with them all, was they did not depict the suspense Hijacked Legacy incorporates.

I considered including the view of a sniper’s scope projected into each of the scenes. That didn’t work, and besides, we had used that visual for Cabin Fever. Instead, I began playing with incorporating a camouflaged individual with a rifle. I figured the next weekend my neighbors were up at camp, and I’d get one of them to pose in camo with a rifle in various locations: on the road and in the woods. Except they weren’t able to come north that weekend, and we were soon heading south,

I’d read about people using AI tools to create images and gave it a whirl. Using Google’s Bard (now Gemini) to create some alternatives, I tried describing the scene as well as I could: a camo-covered sniper with a rifle hiding behind a deciduous tree with its leaves off, and so on and so forth. Somehow, I never got Bard to understand the concept of “hiding behind a tree.” This is as close as I got:

Good enough to play with. First, I eliminated everything but the sniper and then merged him with my photo of the snowy woods. Should he face right or left? How much of him should I show? I came up with a few concepts, passed them by my Readers Group and learned that when they viewed a thumbnail, they knew something was there, just not what. I was also playing with fonts and colors given the neutral background and had mixed comments about those alternatives as well. Shucks.

So, I noodled, and noodled some more, and after many tries, I settled on the final version. In it, I leaned the sniper forward to make him clearer and display the barrel of the rifle. I modified the font and tucked the Legacy tight to the Hijacked, giving the visual impression that Hijacked (with its H and J) is grabbing the Legacy—visually, stealing it.

I hope readers find it intriguing enough to check out. The paperbacks are at the printer and pre-orders are available for both ebooks and paperbacks from my website at https://jamesmjackson.com/ 

* * * * *

James M. Jackson authors the Seamus McCree series. Full of mystery and suspense, these thrillers explore financial crimes, family relationships, and what happens when they mix. To learn more information about Jim and his books, check out his website, https://jamesmjackson.com. You can sign up for his newsletter (and get to read a free Seamus McCree short story).

Google AI Audiobook Creation: A Review

By James M. Jackson

Several months ago, Google Play announced it was beta testing Artificial Intelligence (AI) narrators for audiobooks. My first publisher did not do audiobooks, and when I became my own publisher, I didn’t have sufficient sales to justify directly paying a quality narrator; nor did I feel it was right to convince someone to spend a lot of time doing the narration on a royalty basis. Whether those were good decisions, only an alternate universe knows.

However, when Google Play made its announcement, I had a catalogue of books available to choose from and picked my novella “Low Tide at Tybee” for the experiment. It’s part of the Seamus McCree series but reads very well as a standalone. And it’s short, so if the experiment failed, I would have wasted less time than if I chose a novel.

AI readers are nothing new. For years, I’ve been using Microsoft Word’s version as part of my editing process. But boy does it produce some howler pronunciations. Google takes the ePub version of your book (which must be available at its store) and does its “magic.” You can choose your narrator based on age, sex, and country (US, British, Australian, Indian for English; Mexico and Spain for Spanish). You also pick a narration speed between 50% and 150% of normal. I chose “Mike,” an “American male, Age 31-45” who uses “medium speed and medium pitch.” I raised the speed to 150%.

When the Mike reads a word like read, which it can pronounce as “reed” or “red” it chooses but gives you an easy method to select the other. It’s easy to go through all the occurrences and pick the right one. Some words it simply mispronounces without giving you an alternative (especially common with place names). Google provides you with two ways to correct the error. If you have a microphone, you can speak the word and Google translates what it hears into its phonetic symbolism. I found that approach to work about 75% of the time; the other quarter, it comes up short.

When pronouncing doesn’t work, you can give it a preferred phonetic spelling, which for me requires ingenuity and trial and error before I come up with the right symbolism. Here is an example. It pronounces the word “leaden” something like “leeden” using the symbols: ˈli:dən. The accent on the first syllable (ˈ) was fine, the second syllable dən was also okay. I asked it how it pronounced the word “led.” It told me ˈlɛd, which I combined with dən to create ˈlɛd:dən.

Mike often messes up abbreviations. Seamus used to work for Criminal Investigations Group, shortened to CIG. Mike pronounced CIG as “cig,” like the first part of cigarette. I did a global search and replace (which is easy) to insert spaces between C and I, and I and G. Mike then pronounced each letter. In another book, Mike didn’t recognize the III in Albert Cunningham III as “the third,” instead stuttering the Is. Another global search and replace inserted spaces and fixed the problem.

Sometimes—and I haven’t figured out the trigger—Mike pronounces the word “I” as “the first.” (Which makes me wonder why he wouldn’t pronounce III as “the third.”) I wrestle Mike into submission by changing that I to eye, which Mike always gets right. Also a glitch, Mike occasionally overrides its preferred pronunciation (or the one I have provided) with something else. I think the word before it may cause the glitch. I haven’t found a fix for that issue other than changing words or word order. And twice, it has refused to pop-up the editor to change a word. The first time the global fix of turning the program off and turning it back on again corrected the issue. As of this writing, I have not gotten the second situation to work, resulting in Mike pronouncing the word content like what might be in a bottom drawer rather than how a cat feels being scratched behind the ears.

Those issues are frustrating and real, but relatively rare. I have also noticed over the months that Google’s AI has learned, and words Mike routinely mispronounced when I first used it for Low Tide at Tybee it gets correct while creating audiobooks for the later novels.

The area where AI suffers is inflection. A human narrator would take a sentence like, “She squinted and drew out her reply, ‘Really.’” The reader elongates the pronunciation of “really.” Not Mike. Sometimes with sentences that end in question marks Mike lilts the question at the end. Other times it speaks the sentence the same way as if it had a period. Sometimes Mike now pauses for ellipses and em-dashes. When I first started, he never paused.

When I first started, it took about three hours of my time for every hour of narration. With Granite Oath, it only required about an hour and a quarter for each hour of narration. Part of that is because I have a list of mispronounced words and can rapidly substitute the correct phonetics. But much of it is because Mike is making many fewer errors—it’s learning, exactly what artificial intelligence is supposed to do.

How good is the result? One member of my Readers Group who listens to lots of audiobooks gave Low Tide at Tybee a test listen. He said he was pleasantly surprised that it was better than some human narrators, but not as good as the best ones. I’m fine with that and have priced my audiobooks to reflect the compromise of somewhat lower quality and significantly lower production costs.

Here are links to the completed audiobooks if you want to pick one to check it out. The novels are $5.99 each and the novellas $2.99 each.

AntFarm    BadPolicy    CabinFever    Doubtful Relations    GraniteOath (pre-order)

Furthermore(novella)    Low Tideat Tybee (novella)

So, what do you think?

* * * * *

James M. Jackson authors the Seamus McCree series. Full of mystery and suspense, these thrillers explore financial crimes, family relationships, and what happens when they mix. You can sign up for his newsletter and find more information about Jim and his books at https://jamesmjackson.com.