This is the fourth article in the AI series.

• The first article, Your Wonderful AI Assistant – Sometimes Wrong, Never Unsure, Always Convincing, explains why I’m writing this series and what to expect.
• The second article, All About AI – What It Is, What It Isn’t, and Why It Matters, explains what AI is, where it “came from,” the different kinds of AI, how it’s “trained,” plus examples of how it does and doesn’t work well.
• The third article, AI Assistants – The Good, the Bad, the Ugly and the Unseen, explains how AI tools work (and fail), the different types of AI tools, and when you may encounter them, even when you don’t realize it. This article closes with examples of successfully using AI, along with AI educational resources.

I suggest that you read these articles in publication order, as they build on each other.

AI and Genealogy

Most genealogy vendors use AI in one way or another. Like with everything else, there’s yin and yang to AI and genealogy – but for the most part, sprinkled with a little salt, I love the advances.

Let’s survey the genealogy landscape.

FamilySearch is my go-to for document images, especially now. One of the absolute best things to happen in the genealogy field has been the FamilySearch implementation of full-text searches, which they are rolling out across all of their imaged records. I wrote about the rollout here, and it’s only gotten better in the two years since.

For their latest update, including search hints, watch the video FamilySearch Full-Text Search – Your Golden Path to Ancestral Discovery.

May I live long enough to go back and search for information about each one of my ancestors.

Some of what I’ve already found is, well, shocking and sure does make my ancestors more human. Often, only the primary records like deeds, wills and marriage records have been transcribed, indexed and are online, if that much. Some other records have been published, but you have to know to look in a specific county’s records and have access to the published book. If you were trying to “find your way back in time” to where your ancestor came from, before full-text searching, you were out of luck. Now it’s easy to search across time and locations. Let’s just hope you’re not searching for a common name. Sympathies if you have John Smith.

Is full-text search an application of AI? Absolutely – and it’s wonderful. Is it always accurate? Not by a long shot – but it’s much MUCH better than nothing, and if it’s enough to “get you there,” as genealogists, we can deal with the rest.

Don’t let the pursuit of perfect be the enemy of good!

MyHeritage has been leading the pack with AI for some time now. Initially with their newspaper scans and OCR transcriptions, followed by photo restoration tools, and AI Time Machine. Next, they introduced Tribute Reels, which I wrote about here. Tribute Reels allows you to select photos and then animates them to bring your ancestor to life.

MyHeritage has also recently released Scribe AI, which is now one of my favorites. Scribe AI helps to decipher, translate, transcribe, and interpret old documents, photos, and more. I have not written about Scribe AI yet, but you can read the MyHeritage introduction article here, and try it for yourself here.

As an experiment, I selected a document at random from my saved photos at MyHeritage to see what Scribe AI could tell me. This is a French parish marriage record from 1721. Yes, I submitted something difficult, on purpose, but Scribe AI did not disappoint.

Not only did I receive a historical explanation of what this document is, I received the details, key findings, suggested next steps, a full text transcription in French, PLUS a translation to English, my language of choice.

I have been using ChatGPT for this, but sometimes different AI tools stumble over bad handwriting and such. Mied versus Mius, for example. Now I have two “go-to” tools for translations and transcriptions – both ChatGPT and Scribe AI. Hopefully, if one stumbles over something, the other won’t stumble in the same place.

This is EXACTLY the kind of help that genealogists need.

At RootsTech, MyHeritage announced that they will be introducing GAIA, a genealogy AI assistant, soon. I can’t wait.

Ancestry has been using OCR for their newspaper articles and indexes too, and including items such as wedding announcements and obituaries in your hints. They’ve also implemented AI for LifeStory.

I wanted to see what Ancestry had to say about the LifeStory of William George Estes, who lived in Lynch, Kentucky.

On the next screen, I selected “What might William George have done for entertainment while living here?”

This absolutely cracked me up, because William George Estes was, among other things, a notorious bootlegger. I never heard of him setting foot in a church – not once – nor a library.

Lynch actually didn’t have a library, but it did benefit from the Pack Horse Library Project, where librarians on horses brought the library to remote schools and residents in their saddlebags. Now that would have been both interesting and much more accurate.

Generally, though, this description was relatively accurate. These types of activities were available to people in Lynch, a segregated coal-mining town established and owned by the mining company.

Limited drinking spots? William George counted on that, and it’s exactly how he made his living!

Another recent AI beta tool, “Ideas” has been added and can be found in the upper toolbar of any ancestor’s profile.

I tried “Ideas” for Ollie Florence Bolton because of missing chapters in her story that I’d love to fill.

I was excited. I have already looked for these items, but maybe they are available now, or Ollie’s record can be located. I clicked on “Start researching” for her death certificate.

Was I EVER disappointed. Ancestry, whose primary offering is records, has just told me to search records that IT ALREADY HAS. Ancestry owns Newspapers.com, for Pete’s sake. SEARCH THE DOGGONE RECORDS FOR ME!

As I’ve been using Ancestry over the past few days, I’ve noticed the little Ancestry AI diamond pop up from time to time asking if I’d like help with something. What I wish Ancestry would do is transcribe text in the documents shown, beside the document itself, similar to FamilySearch and MyHeritage.

Across Vendors

One place where we see AI across vendors today is for customer support.

All genealogy vendors and most other businesses now use AI chatbots to answer questions that are asked repeatedly.

AI is the way of the future, and eventually, AI will be able to guide us through records and collections relevant to our ancestors. One day soon, AI will know where to look next and will look there on our behalf, providing an accurate analysis of the results. At least that’s my dream.

Everybody has to crawl before they can walk, so let’s keep moving forward and verifying.

Generating Ancestors

Modifying, enhancing, or even creating photographs of our ancestors is certainly of interest to genealogists. If you haven’t encountered this yet, you will.

I submitted a family picture of the children of Hiram Bauke Ferverda and Evaline Miller to ChatGPT. I asked it to create a likeness of the parents of their adult children, who were standing, and to completely ignore the two seated people. I did not tell ChatGPT that the seated people were Hiram and Evaline.

ChatGPT indicated that it could do as I asked, but cautioned:

• It won’t be historically accurate in a forensic sense—it’s an informed visual estimate based on facial features the siblings share.
• If there are both brothers and sisters, I can usually generate separate likely father and mother composites.

Here’s the output.

Notice that ChatGPT “cleaned up” the images of the children. That means it actually recreated them behind the scenes.

Here’s the second image that is different yet.

Here’s the original for comparison.

Here’s the question we need to ask ourselves and answer.

As ChatGPT cautioned, the images aren’t accurate – and I must add, they are especially inaccurate after ChatGPT “fixed” the original.

Should genealogists do this and attach reconstituted AI images to their trees?

I don’t think so, as they fall into the same category as supposed photos of Revolutionary War soldiers who died in 1800, decades before cameras existed. At least, in that circumstance, with a little bit of applied knowledge and logic, the discrepancy quickly becomes clear.

If an AI-generated image is attached, it clearly needs to be marked “AI-generated” across the photo in a prominent location, larger than in my example, such that the watermark or label can’t easily be removed as the photo is copied from tree to tree in the future.

AI genealogy expert, Stephen Little points to the Coalition for Responsible AI in Genealogy‘s statement on “Protecting Trust in Historical Images” which suggests that we “always label, always cite, and use for illustration not evidence.” It further states that images generated by AI should not be used as historical documents.

Here’s the bottom line. If you put any of these images in your trees, anyplace other than on your own computer (assuming you don’t sync with an online tree), you’ve released it into the wild because it will spread, often as the result of “hints.” Other people will copy and save it, and it will be represented as the real McCoy, either intentionally or otherwise.

We all need to be hypervigilant when evaluating photos, now and in the future.

Custom GPTs and Experts

Some people write “Expert GPTs”. Think of AI, generally, as a general practitioner. An expert GPT is written by someone who either is an expert in that field or thinks they are.

Like everything else with AI, remain cognizant of your surroundings. I mention this because I’m aware of an instance where someone who was very excited and well-intentioned about a topic wrote a GPT that was inaccurate – and if you trusted that GPT as an expert, you would be wrong and unintentionally spreading misinformation too.

If you find an expert GPT, validate that the author actually IS an expert in the topic at hand.

There are several custom GPTs available under the “More” button in the ChatGPT sidebar. You can search by topic or keyword or scroll and view by category.

I searched for Y-GPT, about Y-DNA, because I know positively that Dave Vance is an expert and that the answers his Y-GPT provides can be trusted.

Scroll to the bottom and just click “Start Chat.”

In fact, Dave has pre-loaded his GPT with the most common questions.

Dave’s Y-GPT is a great resource, and I recommend it, but all GPTs are not created equal. Fortunately, there are experts like Dave who freely share their knowledge. Just because someone writes a GPT or THINKS they’re an expert does not make them one. Know your author and their level of expertise before you rely on anything.

Special Project – How I’m Using AI Right Now

I want to share a special project I’m currently working on and how I’m using AI to assist. Actually, AI has saved my bacon.

I’m planning to visit my ancestors’ locations in Massachusetts this year. When I began preparing, I discovered that I have 100 ancestors who hail from 45 different Massachusetts towns. Nobody seemed to stay put. Various reasons existed for moving on. More or better land, marrying someone from another town, religious tiffs within the church, change of religion, or simply the lure of the next frontier.

I’ve compiled research into a document for each individual that includes dates and locations for the following facts, when known:

• Birth
• Marriage
• Childrens’ births and marriages
• Land grants and ownership, including purchases, sales, and inheritance
• Military service
• Civic service or appointments, such as selectmen or jury members
• Church membership
• Wills and estates, particularly land distribution
• Death
• Burial
• Anything else that is location-specific, unique, or particularly interesting, such as if they helped to found a town, a church, or built a fort

If you’ve read any of my 52 Ancestors articles, you’re aware that I enjoy visiting where my ancestors lived, and “being” with them where the events in their lives took place.

That meant that after I compiled the information about each ancestor, I had to reorganize it by town because that’s how the trip would be laid out. Furthermore, in each location, I wanted to find my ancestor’s land, today. This recent article about William and Nehemiah Hunt, who owned Punkatasset Hill, and its connection to the Revolutionary War is a good example.

At this point, I have more ancestors whose land I need to find than I have time to complete the detailed research – and I was beginning to feel incredibly discouraged.

I Accidentally Turned to Google AI

When I say accidentally, I mean I never realized just how useful Google’s AI could be until I submitted William Hunt’s name, birth and death dates, and location to see what other resources Google’s search engine might turn up. When reading the AI overview, I realized that Google had found resources that I had not and certainly would not in the time I have available to plan this trip.

Then, I switched to AI mode, and the resource improved even more.

Incredible!

When using AI tools, we need to take into consideration the strengths of the tool we are using or considering and the resources it has available. In this case, aside from being a search engine and finding information, Google’s specialty is maps, locations, place names, establishments like restaurants, hotels, historical societies, museums and more – including customer reviews.

While I was preparing for the upcoming trip, I stumbled across information that my ancestors, William Hunt and his son, Nehemiah Hunt, both owned Punkatasset Hill, the hill that the men marched down in Concord to take the Old North Bridge in 1775. That infamous “shot heard round the world.”

Google found those resources for me, and their AI tool linked them together. I spent about two days researching this one ancestral place and family for the article – working with Google resources and using each link provided to VERIFY accuracy. You will find that sometimes the resource links don’t contain what AI said they do.

I hurriedly put together William Hunt (c1604-1667), Punkatasset Hill, and the Revolution for America’s 250^th.

I need to make it perfectly clear at this point that AI did NOT write that article or this article for me. I write all of my own articles. Yes, I know AI could write faster and I could publish more, but that’s not my goal or my work. This is my work. Google’s AI was my research assistant, feeding me places and resources to check based on dozens of individual prompts that I fed it in the form of questions.

As you can see in the article, I was able to locate the actual land and home of Nehemiah Hunt, and it was Google that provided the current address gleaned through the video posted by a realtor. Otherwise, I have no idea how I would ever have found that, especially since the local museum/genealogy society is closed on the day I’ll be there.

I was then able to verify that location using other means, including comparing an old photo with a satellite image.

Caution: AI was incorrect more than once, and had I not verified, I would have been off on a wild goose chase, thinking I was catching my ancestors. But without AI and prompting it in different ways, I would never have found this at all.

The Process

I’m going to step through a prompt series for my ancestor Gyles (also spelled Giles) Hopkins, Mayflower passenger and son of another passenger, Stephen Hopkins. I’ll add my own commentary as we go.

I knew that my ancestor, Gyles (or Giles) Hopkins, obtained land in Brewster, Massachusetts. Here’s information I already had before I engaged Google’s AI.

• Stephen Hopkins, in his will dated June 6, 1644, made Caleb Hopkins, his younger son, his ‘heir apparent,’ and consequently the whole of his large real estate holding was inherited by Caleb after Stephen died in July or August of 1644. Soon after his father’s death, Caleb gave a very large tract of land located in what is now Brewster to Gyles Hopkins, his older and only surviving brother.
• As early as 1653, Gyles Hopkins owned the 8th lot in a tract in what is now West Brewster, east of Quivet, which he sold on Nov. 9, 1666, to John Wing of Yarmouth in consideration of a mare, colt and other land.

I wanted to know specifically where this land was located and whether Gyles ever lived there.

My first question to Google was, “Where was Giles Hopkins’ land in Brewster, MA, and where is that today?”

Tip: Be very careful when Googling to include identifying information about the person you’re asking about. For example, “Where was the land of Gyles, also spelled Giles, Hopkins who lived between <date> and <date>…” would have been a better way to ask the question to ensure that I received what I wanted. There could be multiple men with the same name.

Based on what I knew about other aspects of Gyles’ life, I suspected that this land was held speculatively, as an investment, because I knew he had lived in Eastham.

So, I was surprised to see that a home had been built there.

But read carefully – it says, “A historical Hopkins home was later built in this area.” It doesn’t say HIS home, but still.

Plus, there’s a Hopkins Homestead and Bakery? Seriously? You mean, I can visit the land he owned, sit there, and have coffee? Maybe? And who built that house? Is it the original house? So many questions…I had to stop myself from immediately going down that rabbit hole. And I wasn’t even in “AI Mode” yet.

Ok, let’s switch to AI Mode and see what it says.

Now we have the name of Snowy Owl Coffee Roasters too, which says it’s a homestead site, but not necessarily “the” homestead site.

And it looks like Gyles’ (or Giles’) son, Stephen, lived in this location. His grandson apparently was also named Stephen.

Google in AI Mode almost always asks if you’d like additional information – and it’s like crack for genealogists – almost impossible to say no. Keep an eye on the sources on the right, because they change with each question or answer, and you may need to click to “show all.”

Tip: Read the sources; do NOT presume that the verbiage used by AI is verbatim in the sources or that it’s there at all.

Next, I asked to see the bakery on a map. You can tell that I’m considering stopping during my visit, right?

Google just backtracked a bit, saying that the Snowy Owl had moved and that the Hopkins House Bakery sits directly on the land granted in the 1640s.

Hot diggity!!!

Word of caution: Verify addresses and locations on Google Maps directly, as I have found discrepancies.

Keep reading, because there’s more!

Wait? What? There’s a First Encounter Beach monument? I had no idea.

Google asked if I wanted to know more. Yes, of COURSE I want to know more about that.

I already know where Gyles is buried, and I’ll get to that when I’m not focused on his land in Brewster.

Ok, I’m definitely visiting the First Encounter site, but where is it exactly?

Google asked me if I wanted driving directions. I asked for 3 locations to be included: the Hopkins House, the Cove Burying Ground in Eastham, where Gyles is buried, and the monument site.

Google must be experiencing “fatigue” because it can’t seem to retrieve everything from its memory. I’ve had this issue before. I know that Google can easily show multiple locations and stops on maps, but this time, it didn’t show “the map below” that it said was there.

Tip: When I encounter this issue, I close Google and sometimes the browser altogether, then reopen Google in AI Mode.

You can then select the conversation you were just engaged in from the AI history menu. You can also search for keywords and view your most recent AI Mode converations. How convenient is that?!

I noticed an “Upgrade” button appear in the top bar and wondered if that has anything to do with the “fatigue” Google’s AI was experiencing.

So, I clicked on the link to see if this could have been the problem with the fatigue that failed to produce the map. It wasn’t the problem, but the information provided was still useful.

I suspect that if I upgraded to either Plus or Pro, which is very reasonably priced, I might experience fewer display issues. I figured that I might have to do that to complete this time-critical project – but I didn’t.

Instead, I just clicked on the embedded link to the map, and saw that the route had been nicely mapped out.

Google even added a few local features that I hadn’t requested, which I found helpful. I should also note, though, that it missed an important landmark – so don’t get overconfident in its abilities and answers, and always verify.

I didn’t realize that the Cove Burying Ground was just a few feet away from the Eagle Wing Inn. I hadn’t asked for that information. I’m not complaining because I’m going to need a place to stay and I can click on that location and see important information like customer reviews.

Unfortunately, by the time I finished the planning for this project, this Inn was full, but Google recommended other locations based on clustering my visit locations so that I only had to move every few days – establishing a central hub in an area.

Often, several ancestors lived in a specific town, like both Plymouth and Eastham. I compiled my information for each individual location, even if they were adjacent and everything I knew about all ancestors in that location, into a PDF file. Providing AI with too much in a single prompt results in less, more diluted, and less reliable information.

For each town, I uploaded the PDF file and asked Google AI how to find each piece of land, where it is located today, how much time to allocate for each stop, the best order, and even for information on places I’ve missed that would be of interest.

In one location, Google suggested a harbor cruise, complete with a contact phone number and website, to view the land of my ancestor because it’s on Saquish Head, which is entirely privately owned and access-restricted.

In another instance, Google identified a home built by my ancestor that still stands. It has been moved from the original location, which is why I didn’t find it.

I did a LOT of prep work and provided summary information for each ancestor, including location, some of which Google helped me gather. The document totalled 551 pages. I asked Google if it could extract every mention of each town from that document, and it replied that its processing limit was between 50 and 100 pages, so the extraction would have to be done by hand. That’s OK, because it gave me a better feel for the landscape and their lives.

After I extracted the ancestors’ information by town, Google helped me immensely by adding information and organizing it into a plan, with directions and recommended time allocation for each place on the itinerary. In one town, Google even found a local Inn located on my ancestors’ land. Unfortunately, it’s booked, but it is open for dinner, so I’ll still be dining with my ancestor – so to speak.

Here were the project steps:

1. Gather information by ancestor using traditional genealogical resources, plus Google AI to add anything I didn’t already have. Note that Google does not gather information from behind paywalled sites, or sites that technologically deflect crawling and indexing information.
2. Create section in Master Ancestor Document for each ancestor.
3. Ask Google to find ancestral land locations today, adding to their page
4. Extract information from each person in the ancestor document and reassemble by town in travel order. When complete, save as PDF file by town.
5. Submit PDF to Google AI and ask for land locations today – verifying what it said before. Prompt for an itinerary for each town, listing each stop, and how long each town’s activities will take. Include any gotchas such as locked cemeteries, museum hours, inaccessible locations (like islands), and accessibility concerns (walking, steep hills, etc.)
6. After each town was complete, create itinerary in table format. Submit to Google AI for verification and suggested lodging in each area. Ask for driving suggestions to avoid high-traffic areas or congested times of day or other concerns or requests – including lunch or dinner suggestions.

This leaves me with the following documents:

• Master Document by Ancestor – One master document that includes every ancestor in this region. The Table of Contents is 5 pages long, and the document eventually totalled 551 pages. This is how I gathered the information before any organization, and before eliminating Connecticut and Maine, which are outside the scope of this trip.
• Master Document by Location – I quickly realized that the ancestor document was already much too long, so I created a second document by location which totalled 531 pages.
• Individual Location Documents – When finished, I divided the Master Location Document into one file for each location. I eliminated locations where there was nothing to see, so I now have 35 individual files, which have been updated individually using AI, so they no longer match the Master Location Document. I will load these to my laptop and also to a location that I can access from my phone so I can reference them as I travel. I would actually like to print them by location to take with me, except 500 pages is a ream of paper and paper is heavy.
• Master Itinerary by Day/Location – Detailing where I’m staying, confirmation and phone numbers, which town I’m visiting which day, and anything else I need “at a glance.”

What will I need to do after this trip? I’ll need to combine the information from the Master Document by Ancestor with information in the Individual Location Documents that pertains to them. I’ll add photos, of course, and new information gleaned during the trip. I will compile this by ancestor in their own personal file. Will AI be able to help with this step? I don’t know. We’ll see.

And yes, in case you’re wondering, you’ll be reading these as 52 Ancestors stories for years to come!

My intention in this exercize was to illustrate how AI can help locate important geographical features, like ancestral land, for instance, help you organize your data and even plan trips. I almost always find more than I was asking for. I’ve found that Google’s ability to read historical and local history websites is extremely beneficial.

I feared that I might not be ready for this trip in time to make reservations, but thanks to Google AI, I am – and I’ve discovered things about my ancestors I would not have found in time. The worst thing possible is to make a critical discovery AFTER you’ve visited and don’t have the opportunity to return.

Advanced Tools and Bleeding Edge

This series has been focused on the “normie” user, not someone pushing the frontiers of the AI envelope – although we certainly need those pioneers to help sort out how to use these tools safely for genealogy and when to stay away.

When attempting more complex tasks using AI, I’ve experienced spectacular failures. But yesterday’s failures are today’s or tomorrow’s successes. The tools literally get better every day, and a week or month in AI is like a year anywhere else.

The tools from the various AI vendors leapfrog each other, too, and each tool has a unique learning curve.

Recently, Anthropic’s Claude introduced an extremely powerful tool called Fable 5 that was free for the first several days, then available via the purchase of prepaid usage credits or tokens. That’s marketing genius, because it encouraged what I would consider “complex genealogy users” to try it. People who had projects either in mind or in the wings where other models had failed.

Specifically, Fable 5 is focused on what I would call “long horizon” jobs or tasks, such as, “Here are 300 Acadian documents. Please translate and transcribe them, then reconstruct the relationships of the people in these documents.” Or, “Here are 550 pages of ancestors, please reorganize them by town, and within town, by year.” Of course, the construction of the prompt has a lot to do with the degree of success of any query.

Anthropic says it can sustain multi-stage projects over hours or even days, planning work, checking its own output, delegating to sub-agents in agent frameworks, and producing deliverables with much less supervision than previous models.

Some people reported amazing successes with multiple files and hundreds of pages being read, processed, correlated, and reported, mostly accurately. Others had a completely different experience.

I have not been working with these new tools, but Mark Thompson and Steve Little have been. Not only did I not have a massive project ready and waiting, as I had already completed by Massachusetts town extraction, I also needed to have time to VERIFY something that large, because as tempting as this is, AI can be and is often wrong.

You can read about Mark turning Fable 5 loose on his archiving project, here.

Catch up with Steve’s interesting experiments, troubleshooting, analysis and guidance here and here.

The AI landscape is shifting away from asking “Which AI is best?” to asking “How do I assemble the best team of AI specialists and effectively use them?”

Interestingly, these advanced agents are now being used both independently and together to answer complex questions and analyze projects. This is known informally as cross-model verification. Sometimes the models agree, and sometimes they disagree, presenting different “evidence” and debate positions for their conclusions. Ultimately, you are the final arbiter.

Just as genealogists don’t rely on a single record, increasingly they shouldn’t rely on a single AI assistant either. A census, will, land grant, deed, and church record each contribute different evidence, all of which needs to be considered together. Likewise, different AI models have different strengths, weaknesses, and biases. Comparing their answers is often the digital equivalent of seeking independent corroborating evidence.

You don’t need to work on the bleeding edge. Just tuck this away for a future project after these tools are more reliable and stable.

The Take-Away

The takeaway from this is that AI is a force to be reckoned with.

Vendors are using AI, sometimes when we know it, and other times behind the scenes to process text, for example. The net-net of this is absolutely positive.

AI usage will increase over time – and that’s a good thing since we can verify the output. AI can accomplish in seconds, minutes or hours what would take us hours, days, months or maybe never.

As always, AI becomes dangerous when we don’t verify, don’t understand that we need to, don’t know how to confirm the results or we trust “experts” who aren’t.

What Do You Want?

I’m excited for the future and I have my own personal list of how I hope the vendors will integrate AI with tools they provide.

My list includes, but is not limited to:

• AI records transcription at Ancestry, similar to FamilySearch.
• At FamilySearch, full-text records by name and location “suggested” as hints, by individual, along with a multiple-record summary of evidence. Over time, this could go a VERY long way in resolving the “wrong spouse” or “wrong child” connected issues that persist.
• AI “DNA plus tree matching” WITH common ancestors, meaning providing a list of people on your DNA match list who match you and each other and have common ancestors in their tree that you have. Both Ancestry and MyHeritage provide part of this in different ways today, but not all of it.
• AI “DNA plus tree matching” WITHOUT a common ancestor in YOUR tree. This provides a means to potentially identify a brick-walled ancestor by identifying the ancestors of the people whose DNA matches yours and each other, but who do not share a common ancestor in your tree. However, they DO share ancestors with each other, which could be critically important to your research. Yes, I know this would need to be fleshed out – but it’s doable.
• AI records searching. Essentially, what Ancestry suggests as next steps, but to actually DO THE SEARCH on your behalf and provide those results. Go a step further and indicate where to find the records if they aren’t on the platform. Message you if those records ever become available there, or the person has been revealed in existing records. For example, if death certificates are now available but weren’t previously – or the writing was so horrendous that the indexing didn’t pick it up before, but the person has been identified now – let me know.
• At FamilySearch, something similar to relatives at RootsTech, but year-round so that cousins can contact each other, collaborate and make breakthroughs together.
• Broad yet restricted AI searching, such as “provide me with a list of all transactions that include anyone with the surname of Estes or Eastes between 1700 and 1720 in Halifax County, Virginia.” Or, “Provide me with a list of all Estes men who live in Tennessee but who were born in Virginia in the 1850-1880 censuses.” Or, “Provide me with a list of all Estes men who are listed on the 1840 census as being pensioners or having served in the Revolutionary War.”
• At WikiTree, FamilySearch, GEDmatch, Geni and other public sharing platforms,  identify and provide a way to contact individuals who qualify to test for the Y-DNA (male direct paternal male-to-male line) or mitochondrial DNA (everyone’s direct matrilineal mother-to-mother line) of your common ancestors. Include an avenue to offer a testing scholarship at FamilyTreeDNA. There’s so much we could learn!!!

I’m sure you can add to this list. Specifically, what would you like to see AI be able to do for you, either within or outside of a vendor’s platform?

The sky’s the limit, so dream big!

_____________________________________________________________

Share the Love!

You’re always welcome to forward articles or links to friends and share on social media.

Subscribe!

If you haven’t already subscribed, it’s free. You’ll receive an e-mail whenever I publish by clicking the “follow” button at the top of the main blog page, here.

Help Keep This Blog Free

I receive a small commission when you click a vendor link in my articles and purchase that item. This does NOT increase your price but helps me keep the lights on and this informational blog free for everyone. Please click on the affiliate links in the articles or to the vendors below if you are purchasing products or DNA testing.

Thank you so much.

DNA Purchases and Free Uploads

• FamilyTreeDNA – Y-DNA, mitochondrial and autosomal DNA testing
  • FamilyTreeDNA – Y-DNA tests
  • FamilyTreeDNA – mitochondrial DNA test
  • FamilyTreeDNA – Family Finder autosomal test
• MyHeritage DNA – Autosomal DNA test
• AncestryDNA – Autosomal DNA test
• AncestryDNA Plus Traits
• 23andMe Ancestry – Autosomal DNA only, no Health
• 23andMe Ancestry Plus Health

Genealogy Products and Services

• MyHeritage Subscription with Free Trial
• Legacy Family Tree Webinars – Genealogy and DNA classes, subscription-based, some free
• Legacy Family Tree Software – Genealogy software for your computer
• OldNews – Old Newspapers with links to save to MyHeritage trees
• MyHeritage Omni comprehensive “everything included” subscription plan
• Newspapers.com – Search newspapers for your ancestors
• NewspaperArchive – Search different newspapers for your ancestors

My Books

• DNA for Native American Genealogy – by Roberta Estes, for those ordering the e-book from anyplace, or paperback within the United States
• DNA for Native American Genealogy – for those ordering the paperback outside the US
• The Complete Guide to FamilyTreeDNA – Y-DNA, Mitochondrial, Autosomal and X-DNA – for those ordering the e-book from anyplace, or paperback within the United States
• The Complete Guide to FamilyTreeDNA – Y-DNA, Mitochondrial, Autosomal and X-DNA for those ordering the paperback from outside the US

Genealogy Books

• Genealogical.com – Lots of wonderful genealogy research books
• American Ancestors – Wonderful selection of genealogy books

Genealogy Research

• Legacy Tree Genealogists – Professional genealogy research