I have not used the high-end Artificial Intelligence tools that require a subscription, but I want to stay informed about them and how they can enhance, improve and amplify genealogy and family history work.
Recently, I asked OpenAI ChatGPT5 (the free version) to "Please describe complex use cases that ChatGPT Work could perform for genealogists and family historians."
(AI Google Gemini Notebook Inbfographic - Complex Genealogy Use Cases for ChatGPT Work)
The ChatGPT5 response was (split into two parts - this is Part 2, Part 1 is here) is:
The big difference is that ordinary ChatGPT answers a genealogy question; ChatGPT Work can take on a genealogy project—researching, navigating websites, handling files, comparing evidence, producing a finished artifact, and continuing through many steps.
OpenAI describes Work as an agent that can work across apps, files, and the web and stay with complex projects for hours.
For a serious family historian, that opens up some pretty interesting possibilities.
11. Source-citation auditing
You could say:
“Audit the citations in this collection of genealogical sketches. Identify claims that have no citation, citations that don't appear to support the claim, duplicate sources, incomplete citations, and sources that should be upgraded to original records.”
That could be followed by:
“Produce a corrected citation list in Evidence Explained style.”
For a large body of genealogy writing, that's potentially a tremendous time saver.
12. Find contradictions in a genealogy database
This is another area where an agent could shine.
For example:
“Find everyone in this family tree who was supposedly born before one or both parents, married after their reported death, lived in two distant places simultaneously, or had an implausible age at marriage.”
Then:
“For every anomaly, locate the underlying records and determine whether the problem is a transcription error, an incorrect person, an incorrect date, or an unresolved conflict.”
That's essentially quality-control testing for a genealogy database.
13. Build a migration study
Suppose you want to understand the Carringer migration:
Pennsylvania → Iowa
You could ask:
“Trace every Carringer household in my tree that migrated from Pennsylvania to Iowa between 1840 and 1880. Determine approximate migration dates, counties, occupations, land purchases and known associates. Identify whether groups of related families migrated together.”
Then have Work produce:
- a timeline
- a table
- a map
- a narrative
- source citations
That becomes a genuine family migration study, not merely a genealogy search.
14. Military-service reconstruction
Your father's Navy experience is a perfect example.
You could give Work:
- DD-214
- muster rolls
- enlistment information
- ship histories
- Navy training documents
- photographs
- your existing notes
and say:
“Reconstruct Frederick W. Seaver Jr.'s Navy service from enlistment through his arrival aboard USS Halford. Identify what is documented, what is strongly supported by Navy procedures, and what remains uncertain.”
Then:
“Research what a newly enlisted Navy man at NTC San Diego in September 1944 would probably have experienced during his first week, using contemporary Navy documentation.”
The result could become a historically grounded family memoir chapter, with the documented portion kept distinct from reconstructed context.
15. Photograph detective work
Give Work a folder of family photographs and ask it to:
“Create an inventory of these photographs. Identify photographs that appear to depict the same person, compare backgrounds, clothing, facial characteristics and photographic formats, estimate approximate dates, and identify photographs that might belong to the same event.”
Then:
“Cross-reference the apparent dates with my family tree and newspaper research.”
That could be especially useful with your large digitized photo collection.
16. Cemetery research
Imagine:
“Research everyone buried in Columbus City Cemetery with the surname Carringer or identified as a relative. Build a cemetery family network. Compare gravestones with my database and identify discrepancies in dates, spouses and relationships.”
Then:
“Find obituaries for everyone for whom an obituary appears to exist and compare the obituary information with the gravestone.”
That's a multi-source reconciliation project.
17. Create a genealogy book from raw material
This is where Work moves from research assistant to production assistant.
Give it:
- four-generation pedigree
- biographies
- photographs
- documents
- family stories
- timelines
- citations
and say:
“Create a 4-generation Seaver family history. Organize it into chapters by generation. Insert photographs in appropriate places, standardize names and dates, create a table of contents, create source notes, identify missing photographs, and produce a print-ready manuscript with a name index.”
Work can create finished documents, spreadsheets and presentations from the material it works with.
That is considerably more ambitious than “write me a biography.”
18. Turn your genealogy into a website
You could give Work:
“Create a family-history website for the Seaver family using these biographies, photographs, timelines and pedigree information.”
ChatGPT Work can create interactive Sites/web experiences as part of its workflow.
You could then have it maintain the site:
“Add the newly completed Isaac Seaver biography and update the descendant timeline.”
That starts looking like a living family-history archive.
19. Turn a research project into a presentation
For a genealogy society presentation:
“Take this research project and create a 45-minute presentation explaining how the brick wall was investigated. Include the evidence timeline, research methodology, documents, maps, DNA evidence, unresolved questions and lessons for other genealogists.”
Work can produce the slides and supporting documents rather than merely giving you an outline.
20. Your weekly Genea-Musings assistant
This may be the most practical one for you.
You could establish a recurring workflow:
- Every Monday: Find an interesting person in my research database who has a useful document that I haven't written about recently. Prepare an Amanuensis Monday candidate.
- Every Tuesday: Find a Seaver-related newspaper item suitable for "Seavers in the News."
- Every Wednesday: Identify an interesting family photograph for Wordless Wednesday.
- Every Thursday: Identify an interesting historical record suitable for Treasure Chest Thursday.
- Every Saturday: Define a unique, but relatively simple, genealogy research challenge for "Saturday Night Genealogy Fun."
Work's scheduled tasks can run recurring or event-triggered workflows and can monitor sources for changes.
That moves AI from something you ask questions of to something that helps maintain your genealogy workflow.
21. The really ambitious one: “Genealogy Research Manager”
Here's where I'd push it.
Give Work a research project with:
- Research question
- Known facts
- Existing evidence
- Family tree
- Research log
- Research standards
- Repositories to search
- Citation requirements
Then tell it:
“Act as my genealogy research manager. Work through this problem systematically. Do not repeat searches already completed. Maintain a research log. Record negative evidence. Distinguish facts, hypotheses and assumptions. When you find promising evidence, pursue the associated record sets. Stop and ask me when a decision requires my judgment. At the end of each research session, update the research log and give me the next three recommended actions.”
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
That is, in my view, the conceptual leap with Work.
ChatGPT says: “Here's what I found.”
Work says: “I investigated the problem, searched the sources, organized the evidence, documented what failed, created the research report, and I'm ready for you to review the next decision.”
And that distinction matters because genealogy contains an enormous amount of repetitive, multi-step, clerical work surrounding the actual historical reasoning.
The important limitation
I wouldn't treat Work as an autonomous genealogist whose conclusions you simply accept. Websites can block automated agents, records can be misindexed, handwriting can be misread, and genealogy's biggest danger remains turning a plausible person into the right person without sufficient evidence. OpenAI specifically recommends reviewing sources and important results, and Work pauses for sign-ins and consequential actions when appropriate.
The sweet spot is therefore:
- You = genealogist, historian, final judge
- Work = tireless research assistant + browser operator + file analyst + data clerk + writer + production assistant
For someone with your volume of accumulated research, I'd actually divide your Work use into five “genealogy agents”:
Brick Wall Investigator — solves difficult ancestry problems.
Evidence Auditor — checks sources, citations and contradictions.
Record Miner — newspapers, land, probate, military and census.
Ancestor Storyteller — turns documented research into biographies and historical narratives.
Family History Publisher — turns the accumulated material into books, presentations, websites and blog material.
That combination could be much more consequential for a serious family historian than simply having a smarter chatbot.
In every complex genealogy use case on this list, there are questions like "how can this be done?" and "what information, and in what format, do I need to provide?"
A big challenge for many of these complex use cases is that records needed are behind a commercial website paywall (e.g., Ancestry, MyHeritage, Findmypast, American Ancestors). Some of those websites have explicit restrictions on any AI tool being used to use the website to find records and information. They require the subscription user to do the searches.
It is evident that ChatGPT Work, like Claude CoWork, relies on an ongoing dialogue between the user and the AI tool for many use cases, and requires access to the user's digital files that have the records to be analyzed.
Some of these twenty-one use cases can be performed using the free version of ChatGPT; for instance, #9, "Automatically turning research into sourced-based biographies."
Other researchers have performed some of these use cases using ChatGPT Work, Claude CoWork, and other AI Tools.
I may write blog posts about some of these complex use cases to answer those questions about them.
Note that ChatGPT5 uses my own research names, dates and places as examples for these use cases.
Links to my blog posts about using Artificial Intelligence are on my Randy's AI and Genealogy page. Links to AI information and articles about Artificial Intelligence in Genealogy by other genealogists are on my AI and Genealogy Compendium page.
The URL for this post is:
Copyright (c) 2026, Randall J. Seaver
Please comment on this post on the website by clicking the URL above and then the "Comments" link at the bottom of each post. Share it on X, Facebook, or Pinterest using the icons below. Or contact me by email at randy.seaver@gmail.com. Note that all comments are moderated, and may not appear immediately.
Subscribe to receive a free daily email from Genea-Musings using www.Blogtrottr.com.
No comments:
Post a Comment