Genealogist, author, and CCOGS Lineage Societies Chair Linda Collins reflects on research, storytelling, forensic genealogy, and the growing role of AI in family history.
Many genealogists are curious about artificial intelligence, but aren’t quite sure where it fits into their research. Can AI help solve long-standing brick walls? Is it worth learning? And how can researchers use it responsibly?
Those are some of the questions explored during our September meeting with Linda Collins, Chair of the CCOGS Lineage Societies Committee. A genealogist with more than a decade of research experience, Linda is also the author of the historical novels Dreams of Revolution and Ship of Despair. Her interests have taken her from traditional family history research to forensic genealogy and even the use of AI as a research and writing tool.
In the interview below, Linda discusses her first experiences with ChatGPT, shares lessons she has learned along the way, and explains how artificial intelligence can become another useful tool in a genealogist’s toolbox.
Genealogy is built on careful research, while artificial intelligence is still a new and sometimes controversial technology. What inspired you to experiment with ChatGPT in your genealogy work? Were you skeptical at first, and if so, what did you discover that changed your perspective?
I first used ChatGPT to give me information I needed for my shipwreck novel. The men were drowning and starving, and I didn’t know how to describe how they would look and feel. AI gave me lots of good information about those conditions. Then I experimented with improving my writing with it. Later, I saw videos on using it for genealogy, then tried ideas for questions I had, and the presentation was the result.
As a genealogist, you have spent years developing systems for organizing records, photographs, and research notes. AI can quickly uncover new leads and unexpected connections. How do you balance exploring those possibilities with staying focused on your original research question, and what best practices would you recommend to other family historians?
As far as exploring new AI leads, I haven’t used its suggestions as much as I would like to due to lack of time. I looked at some of the places ChatGPT suggested for finding civil war pictures, but so far I haven’t been successful. I especially liked ChatGPT’s observations on my mystery man, George Grubaugh, and its theory that I’m finding facts on two different men. George has been a problem for many years! As far as best practices, I try to keep a research log of the places I’ve searched. That way I hopefully don’t cover the same ground when searching later.
The AI landscape seems to change almost daily, with new tools and features appearing all the time. As someone who has explored AI for genealogy, forensic research, and storytelling, what qualities make an AI tool truly useful? Is there a platform you return to most often, and what advice would you give researchers trying to decide between a free account and a paid subscription?
AI is most effective when you have a clear, narrow question to ask it. You can always expand your question depending on its answers, and it often gives good suggestions for using other ideas and methods to search. ChatGPT free version is the only one I have ever used. There were only a few times when it balked at answering too many questions at one time, but that was usually not a problem. If it gets uncooperative it tells you when you can come back and ask again, which is usually just a few hours. If you are using it a lot, then a paid subscription might be useful.
Your genealogical work has taken you from tracing family histories to assisting with forensic genealogy projects that help solve real-world mysteries. What inspired you to pursue this path, and what has been the most meaningful or memorable case you’ve worked on? Tell us the story of the project and how genealogical research played a role in finding answers.
I’ve always been interested in crime shows, and when some high-profile cold cases such as the Golden State Killer were solved using investigative genetic genealogy (IGG), I was hooked. I wanted to use my genealogy skills to help solve crimes and identify unidentified human remains. Many victims of crime have gone decades with no resolution, and this is a means to help them. I took an on-line class through Ramapo College in New Jersey about IGG. Unfortunately, the only two databases available for IGG work are Family Tree DNA and GEDmatch. I’ve always used Ancestry.com, but their terms of service prohibit any law enforcement use. That also rules out any Ancestry owned companies such as Find a Grave and newspapers.com.
I had to hone my skills with GEDmatch and FamilySearch to work cases. The class worked cases as a group, and our first actual case was from North Carolina. The basics are: find a DNA match to the unknown person, identify them (sometimes a sticky problem), then build a direct line family tree for the match. Then multiple matches generate multiple family trees and the goal is to find the ancestral couple where those trees match, because your unknown person also has that couple as their ancestors. Then you build the tree down from that couple using ALL the descendants and look for missing people in the right time frame as your unknown.
Our first case was from North Carolina, and we found many, many intermarriages and the trees were a mess. We didn’t solve that one, and the college will conduct some specialized DNA testing on the problem areas. The second case was from Arizona which we solved. The young man we identified had been missing since the 1980s. Supposedly he had an argument with his father, went out to buy beer, and never returned. Only his skull was found the same year with a bullet hole, but he had been unidentified all that time. Actually, his skull had been kicking around the police department as a specimen forever. We were pretty sure we were correct because his family had a Facebook post telling about his disappearance, and all the information matched up. His name has not yet been released yet because the police have to test the DNA of his relatives to make sure it is him. Also, now that they know his identify, they may be able to solve his homicide, so don’t want the killer to know they are working the case again.
Tell us more about your historical fiction novels.
My original historical novel I published is Dreams of Revolution, and the latest one is called Ship of Despair. It’s a true shipwreck story about 35 men traveling from New Bedford, Massachusetts, to Savannah, Georgia, for winter work. Only 5 men survived. I researched everyone’s genealogy, and found more interesting stories which might lead to the next novel. My website is: www.lindacollinsauthor.com but I haven’t kept it very up-to-date. That’s the next project!
Tell us more about your new non-profit.
I formed a non-profit corporation: Collins Forensic Genealogy Center, to work on cold cases, but so far I haven’t gotten any cases. I’m researching cold cases in Mansfield and think I’ll have to make a visit to the coroner to see if there is DNA for them.