A woman walked into Kennisha Thornton’s office a few years ago with news she was excited about.

She had already published one book with Kennisha’s help, so she knew what the work took. Months of it. Drafts, edits, the whole slow climb. This time she had found something better. “Hey, I think I found a way that I can make, like, seven books, like, within a an hour or so,” she said.

Kennisha asked what she meant. Full books? Chapters and everything?

“Yeah, like I, I think I just put in what I want here and it’ll just make the books for me.”

This was before ChatGPT. Kennisha’s reaction was the one you would expect from someone whose entire living depends on original writing. “I don’t know that I would trust that,” she said, “because where’s the information coming from and how do you know it’s safe?”

I bring this up because of where she ended up. The woman who asked that question now uses AI every single day. She uses it to build her talks. She uses it to write video scripts. She uses it to get past the blank page. And when I asked her how she got from there to here, she gave me the clearest answer I have heard from anybody on where the line actually sits.

She started with the smallest possible thing

Kennisha did not begin with books. She began with Canva.

“They had a feature that I would never forget when I learned about it,” she told me. “Like, oh, well I can use this AI feature to help me reword a few things on my graphic.” That was it. A few words on a social media graphic, reworded.

What made it safe was the ownership. “If I’m just asking it to rewrite something I initially wrote, I’m just asking it to give me another way to say it.” The words were already hers. She was asking for a second option, the way you might ask a colleague to look at a sentence you have stared at for too long. Nothing was being invented. Nothing was arriving from a source she could not account for.

From there it grew slowly. She started using it in her design work. Then she started putting small prompts in when she got stuck. “Hey, is there anything else I can add to this? What do you think?” Small asks, one at a time, over a long stretch. “Little by little I began to research and try and see how this could work for me and get more comfortable.”

There is a lesson buried in that sequence that I think most people miss. Every argument about AI online is pitched at the maximum setting. Generate a whole book. Replace your whole team. Automate your whole business. Kennisha’s path was the opposite shape. She found the smallest version of the thing where she still owned every word, tried that, and only widened the circle once she trusted what came back. That is how careful people adopt anything, and it is almost never how AI gets sold.

Almost 20 years of watching what actually reaches people

To understand why she was careful, you have to understand what she does when AI is nowhere in the room.

Kennisha’s work is stories. Not as decoration, as the load bearing part. Plenty of the authors she works with are business leaders who want a book that establishes them as the authority in their field, and they arrive with expertise stacked up and ready to pour onto the page. What she has learned over almost two decades is that the expertise alone does not land.

“I love that we get to share the knowledge that we have,” she said, “but people really do, like, stories stick with us.”

She reached for a comparison I have not stopped thinking about. A children’s picture book has the sweet story and the big beautiful illustrations, and the illustrations tell the story just as much as the words do. Her point was that a business book has the same two tracks available to it. The knowledge is one. The stories are the other. Authors who only run the first track are publishing half a book and wondering why it does not move anyone.

And she is specific about why it works. “It caters to our feelings. It moves our emotions, and that helps us make decisions.” I spent years in sales before I was a marketer, and the line drilled into every salesperson is that people buy on emotion and use data to justify what they already decided. Kennisha arrived at the same conclusion from the publishing side, working with authors instead of buyers. That is two different roads to the same intersection.

She teaches this at Southern Careers Institute, a trade school in Waco, where her students are training for certifications in hands on fields. Her classes are customer service and communication. I asked whether she gets pushback from people who signed up to learn a trade and have no interest in storytelling. Her answer was about reading the room. “If I see that, okay, I’m probably going too far here, I pull it back. We get back on track.” Then she added something that stuck with me: “Asking is also probably lost, that we probably don’t ask if it’s okay to share instead of just assuming everyone wants to hear your story.”

The rule she gives authors

So here is a person who knows exactly what a story is worth, and exactly how much work goes into getting one out of somebody. That is the background against which she draws her line.

The line is ownership of the thinking.

“I think it’s always best if authors use their own original thought and ideas,” she said, “and if we could use something to help us maybe edit or reshape or rewrite, it’s our original work that’s just being edited, and maybe that’s okay.”

That is the rule. The idea starts with you. The shaping is yours. The tool works on material you supplied. What she warned the woman in her office against was the other version, using it “for your entire book, like, without you having any say or thought into what you’re trying to say.”

Watch how that rule applies to her own work now. When she built a presentation about careers in the trades recently, she used AI to put it together. But she went in first: “This is what I want to talk about. These are the points I need to cover. So help me organize and structure this.” Then she went back in after, and added the part no tool could have supplied. She told a story about her father, who worked in the trades for 44 years.

“I could still go in and add my authenticity, my storytelling, and make it still unique and special, but AI can help me keep that structure in place, and make sure that I cover every point that I need to cover.”

Her own summary of the arrangement is the sharpest sentence in the interview. “It is not thinking for me, but it’s just building based on what I’m giving it to build. So I’m essentially still building, but it’s helping me put the pieces together.”

The story about her father was never going to come from a machine. The outline that made sure she got to it was.

The part that will start an argument

Now the harder case, and the reason I wanted this conversation on tape.

Kennisha recently worked with an author who used AI to draft her entire book. Not an outline. Not a stuck chapter. The whole draft.

And Kennisha’s verdict on it was that the book belongs to the author. “Now it was her concept, her idea, and it took time to shape it, to make it what she wanted it to be, but it was still her, it’s her work, her original work and ideas, but she just used it as a tool.”

Plenty of people will disagree with that, and they will have a real case. There are respected voices, including ghostwriters I follow on LinkedIn who have built their whole position on this, who hold that a book drafted by AI is not the author’s book no matter whose concept it started from. That argument deserves a hearing. Drafting is where a lot of thinking actually happens, and a writer who skips it may not know what they think.

What makes Kennisha’s position worth taking seriously is who is holding it. She is the one with almost 20 years and hundreds of authors behind her. She is the one who was suspicious first. She did not arrive at this from enthusiasm about technology. She arrived at it from watching what came back, book after book, and judging whether the person on the cover was still the person on the page.

Her test is not how much of the typing the tool did. It is whether the concept, the shaping, and the judgment stayed with the human. By that test the book passed.

Why she stopped worrying about the accusation

There is one more thing worth carrying out of this conversation, and it is about the criticism rather than the tool.

The charge people throw is that using AI means you did not really do the work. I put a comparison to Kennisha during the interview and she took it further than I did. Think about the executives you have seen in films, walking down a hallway with an assistant beside them taking everything down. Nobody accuses that executive of not doing their job. Nobody says the ideas were the assistant’s. We understand instantly that the thinking belongs to the person talking and the capture belongs to the person writing.

That is the arrangement Kennisha describes. The AI became the assistant with the notebook.

She has been at it long enough now that it has learned her. “My AI probably knows me like the back of its hand,” she said. “I’ve groomed it to know how I think at this point.” That word, groomed, is worth sitting with. It is time she put in. The tool did not arrive knowing her voice. She taught it, in the same slow careful way she started with a few reworded words on a Canva graphic.

What this means for you

If you have been circling AI and not touching it, do what she did rather than what the internet tells you to do.

Find the smallest task where you already own every word. Something you wrote, that you want said another way. Ask for that and only that. Judge what comes back against work you know is yours. Then widen it one notch, and only when the last notch earned your trust.

And when you get to the bigger work, hold her rule. Bring the concept. Bring the shaping. Bring the story about your father, or your first job, or the customer who changed how you do business, because that part exists nowhere else and no tool will hand it to you. Let the machine build the structure around it and make sure you hit every point you meant to hit.

Kennisha speaks at the YouXAI Summit on September 12 in Waco, on The Human Story in an AI World. She is going to walk through how she actually uses it, day to day, as a creator. If you write anything for a living, or you have a book you have been carrying around for years and have never started, that is the session to be in the room for.

Tickets are $79 through August 31. They go to $99 on September 1. Register at youxai.live.

— Fernando Labastida YouXAI Speaker Preview Series

P.S. Her podcast is called Beneath the Surface, and season one is in production now. The whole idea of it is that we meet each other at the surface, at networking events, with “oh, what do you do,” and almost never get past it. She wants the part underneath: what it actually took to get where you are. Worth watching for when it lands.