What if your research refused to stay silent—prodding you, murmuring secrets, invading your dreams, and urging you to keep searching?
In this week’s episode of Flower in the River, I did something a little different. After sharing my 2025 retrospective, The Search Goes On — Coincidence. Clarity. Resolve., I got a lot of responses—more than I expected. So I decided to step back and try to understand why it resonated. I gave the transcript to Google’s NotebookLM—an AI tool gaining traction among genealogists—and let it dissect my work.
The outcome? Two AI research pals, “Eva” and “Max” (NotebookLM’s AI hosts), listen to my last episode and talk back—analyzing the research, the revelations, and the questions that bubbled up. What caught their attention: a demand for accountability and higher standards in documenting historical tragedies.
It’s thoughtful. It’s weird. It’s a little surreal to hear AI reflect my own work back at me. And it sparked insights I never saw coming.
What “Eva” and “Max” Noticed
The AI hosts noticed something I hadn’t put into words: the tug-of-war between repeated “history” and grassroots research. They named my approach “deep listening”—a blend of endless patience, openness to nonlinear clues, and sharp pattern recognition.
They also didn’t shy away from the harder questions. When I found a man inadvertently omitted from George Hilton’s original estimate of victims—they understood immediately what that meant. The earlier number had become a brand. And brands are harder to change than facts.
Then there’s the woman who died seven years after the disaster, but whose family always insisted she got sick right after being rescued and never recovered.
Eva and Max asked the question I’ve been wrestling with: What do we even define as a casualty in a mass tragedy? There are no official standards for the Eastland. And there will never be a final number. All numbers are estimates.
What Emerges
Hearing AI voices analyze your own work is both intriguing and bizarre. It’s like having someone hold up a mirror, except that someone is made of code. They caught things I didn’t consciously realize I was saying. They missed some things, too—but they got more right than I expected.
What emerges from this episode are the themes of clear sourcing, agreed definitions of casualty, and transparency.
History is never really “finished”—and the inquiry must continue.
Flower in the River Year-End Visual Recap (2025)
Buzzsprout, the platform that hosts and distributes my podcast, provides an annual listening recap for podcasters. I also took their year-end recap slide and stitched together my episode posters (all 53 of them) in a fast scroll. It moves quickly—but seeing them together was a reminder of how much ground steady work can cover. A lot of research. A lot of writing. A lot of people whose stories now have a permanent home. Very cool.
If you’ve been along for any part of this ride, thank you. If you’re just joining now, welcome—there’s plenty to catch up on.
