I’ve combined Amy Johnson Crow’s 52 ancestors in 52 weeks challenge, and Steve Little’s The 2025 AI Genealogy Do-Over, to create a unique 52 AI ancestors in 52 weeks party!

52 AI Ancestors in 52 Weeks: Week 9: Family Secrets

Introduction

This week’s discussion is on Family Secrets. My paternal grandmother told me many family history stories – all of them wrong. I loved Grandma, but boy, did she spin them.

Discussion

One of the stories I got was about her maternal grandfather, James HART. When I was a baby genealogist and new to the hunt, I did manage to do the “family interviews” part. Grandma told me that James was killed “in a hunting accident” when she (born 1916) was “a girl.”

To say that I searched for years for his death is an understatement. When my boys were small and I was a Cub Scout den leader, we visited a state park where I asked a Park Ranger about researching hunting deaths from the 1920s. They gave me a phone number but it didn’t get me anywhere.

I didn’t find his death in the vitals indexes during the time period I thought it would have happened.

When newspapers started coming online, I spent many late nights getting bleary-eyed.

Finally one day, after my grandmother and father died in quick succession, I was going through family photos and found one labeled “Bob [my dad] with Joseph and Joe <redacted>.” That surname had cousins on this line so I did an internet search to get the photo back to the family. I connected with a cousin and we spent time discussing family history.

“James?” he said. “Oh, he shot himself.”

Wait, what?

Must have been a hunting accident, right?

Nope.

The newspaper headline is “Despondent Man Ends Life With Shotgun.” He was not in the woods and he very deliberately exited this world.  

Grandma, your family secret is no more.

Depression runs through our family, but until recently no one would discuss it.

How AI is Helping My Research

I asked AI how it can help me identify patterns of depression running in a family. It offered several options:

1. Analyzing Obituaries and Newspaper Reports

AI-powered tools like Newspapers.com and Chronicling America use Optical Character Recognition (OCR) to search for terms like “despondent,” “melancholy,” or “troubled” in historical newspapers, which might indicate depression or suicide.

AI sentiment analysis can review obituaries or death reports to detect language patterns related to mental health struggles.

2. Detecting Mental Health Clues in Family Records

AI-assisted genealogy platforms (like Ancestry and MyHeritage) can analyze cause-of-death records across generations, looking for recurring trends such as “self-inflicted wound,” “nervous breakdown,” or “institutionalized.”

AI handwriting recognition tools can decode personal letters, diaries, and census records to find mentions of mental health conditions.

3. Cross-Referencing Medical and Historical Data

AI can help analyze old medical records and census documents to identify patterns of institutionalization or hospitalization in a family tree.

Tools like Google’s BERT AI model or ChatGPT can scan large volumes of historical text and extract relevant information related to mental health.

4. Family Tree Pattern Recognition

AI tools can analyze relationships and life events to identify stressors that may have led to depression or suicide, such as economic downturns, family deaths, or war experiences.

Predictive modeling could compare your family’s history to broader social trends, revealing external factors that contributed to mental health struggles.

5. Speech-to-Text for Oral Histories

AI transcription tools (like Otter.ai or Whisper) can convert old family interviews into searchable text, helping you detect recurring themes of depression or emotional distress in family stories.

6. Photo Analysis for Hidden Clues

AI-powered facial recognition tools might help analyze historical photos for physical signs of illness or distress.

Some AI models can even compare expressions across multiple generations to spot potential inherited traits related to emotional well-being.

Challenge for Readers

Mental health research may not be as thrilling as chasing an elusive ancestor, but it’s just as important. It’s crucial to have the knowledge in order to combat it.

My recommendation: give your GEDcom to an AI model such as Claude or ChatGPT and ask it to look for words indicating mental health struggles, identify the people, and identify clusters or potential inheritance of such. Add in other factors such as military service and other situational and environmental factors. With that information, we may be able to identify those who are more at risk, and take early steps for treatment.

Summary and Next Steps

I’m so grateful that my father, a young adult in the 1960s, thought it important to be open and honest with me, and impressed on me the value of keeping the lines of communication open, even as we had our own mighty struggles. He worked so hard to fight the secrecy he’d been raised in so that we could do better.

Grandma, the truth is finally known—and it’s for the better.

The photo that broke the brick wall

Disclosure

This post was created by me and refined with AI assistance. While AI helps organize research, the storytelling and discoveries are my own.