How AI Helped Me Rediscover My Own Book and What It Means for Every Writer
- Rhea Wessel
- 7 hours ago
- 3 min read
By Rhea Wessel
Note: Scroll down to listen to the podcast of my book created with NotebookLM
Not long ago, I ran an experiment that became something more powerful than I expected.
Using NotebookLM from Google, I uploaded the PDF of a book I had written and lived with for a long time—something close to my heart, developed over many years, with the help of talented editors and thoughtful collaborators. What happened next was astonishing.
The tool created a podcast version of my book. But this wasn’t just an audiobook or a robotic summary. It was something else—something more intelligent. It found the story arc. It surfaced the quotes that landed best. It revealed, with startling elegance, the emotional thread that carried the narrative across chapters. In one sitting, I found myself laughing—and crying.
What shook me most was not that the AI was accurate—it was that it understood the work at a level I hadn’t anticipated. This was a book that had been written organically.
I didn’t use AI tools during its creation. I didn’t map a rigid structure. The narrative flowed from a deeply human place: intuition, lived experience, and many conversations with editors and readers. The arc was coaxed into shape with human care. Yet here was a machine, retroactively surfacing that architecture with clarity and confidence, as if to say: “Yes, it’s here—and it works.”
And when it doesn’t work? The tool can make that visible, too.
Since then, I’ve used NotebookLM on several projects. The results are consistent—and revealing. If the paper lacks structured thinking, the podcast will sound like a string of unconnected facts. If the insights are shallow, the audio won’t linger in your mind. If the writing rambles, the spoken version leaves you disoriented, asking, “Wait—what did they just say?”
That’s the real magic of this tool. It’s not just about automation. It’s about feedback. Honest feedback. If the content lacks a spine, the auto-generated podcast version will show it. If the ideas are fresh and the structure clear, the output sings.
This, to me, changes the game—especially for anyone working in thought leadership, long-form writing, or even internal corporate communication. You don’t need to wait for a peer review or an editorial board to find out if your piece has legs. You can hear it. You can feel it.
I’ve started thinking of this podcast feature as a kind of “litmus test” for content. If the audio version of your work lands with clarity, if it holds attention and builds meaning moment to moment, then you’re likely working with something strong. If it confuses or overwhelms, the problem is not the tool. The problem is likely in the writing, the structure, the sequencing of thought.
It’s one thing to read your work aloud to yourself (a practice I still recommend). It’s another to hear it interpreted and presented back to you by a machine with no bias. Yes, no bias: NotebookLM relies only on what you give it. It’s not “allowed” to source more widely.
It doesn’t care how hard you worked on a paragraph. It just processes what’s there—and shows you whether or not it makes sense.
This technology isn’t replacing editors, writers, or designers. But it is revealing what’s under the hood—faster, clearer, and often more honestly than we can see ourselves.
So if you’re working on a book, a white paper, or even a slide deck that’s meant to influence others—try this.
Upload it to NotebookLM. Listen to what comes back. Does it move you? Does it make sense? Could someone in your audience follow the logic, remember the key points, feel what you want them to feel?
If the answer is no, that’s not failure. That’s information. That’s insight. That’s an invitation to tighten, clarify, and focus your ideas.
I walked into this experience curious and casual. I came out the other side humbled.
Listen to the podcast here: