RLP 405: Interview with Mark and Steve of the Family History AI Show

Nicole (0s):
This is Research Like a Pro episode 405: Interview with Mark and Steve of the Family History AI Show. Welcome to Research Like a Pro a Genealogy Podcast about taking your research to the next level, hosted by Nicole Dyer and Diana Elder accredited genealogy professional. Diana and Nicole are the mother-daughter team at FamilyLocket.com and the authors of Research Like a Pro A Genealogist’s Guide. With Robin Wirthlin they also co-authored the companion volume, Research Like a Pro with DNA. Join Diana and Nicole as they discuss how to stay organized, make progress in their research and solve difficult cases. Let’s go.

Nicole (42s):
Hi everyone and welcome to Research Like a Pro.

Diana (46s):
Nicole, How are you doing today?

Nicole (48s):
I’m doing really well. How about you? What have you been up to?

Diana (51s):
I’m doing well. I’m working on research planning for the study group and I have to say I have a whole project on my Cynthia Dillard projects. I have so many different things on her in there, past reports, you name it, I’ve got it in my, my Claude project and yesterday I experimented and I said, okay, just write me a hypothesis for this new phase. And it did really good. I was very happy with it.

Nicole (1m 16s):
Nice.

Diana (1m 17s):
So interesting. And I think it helped because I had some past project documents where I had sample hypotheses maybe ’cause I didn’t give it a sample, I gave it my objective and you know what I currently know. But anyway, so interesting. I’m just comparing back years ago when it was not nearly as good. But I think having all that project knowledge was just key. Anyway, I’ve been having fun experimenting with AI and kind of pushing the limits there.

Nicole (1m 43s):
Yeah, that’s awesome. I’ve been trying AI to help me with pedigree triangulation. I opened up 20 tabs of different public member trees all from one Ancestry cluster. I made a custom cluster and then I had Claude in Chrome, which is a browser extension sidebar, I had it analyze all the trees and list all the surnames and find surnames in common among the trees. And then I just chatted with Claude about those tab groups And it was super helpful. I don’t like going through so many tabs and trying to remember things and, and in the past I’ve tried to paste out all the surnames, but it’s a multi-step process And it was so nice to have AI help with that.

Nicole (2m 23s):
And of course it’s an agent so it’s a little bit more risky. And so you have to really learn about the risks and not put any tabs in the tab group from websites you don’t trust. But once you learn about that and gain comfort with it, if you are able to gain comfort with it, I know I’m more on the like try new things end of people. And so it was just really helpful.

Diana (2m 45s):
Okay. So did you have to expand those trees at all or did you just leave them as they were? ’cause you know that final generation then you can expand it or did you just leave it as it was when it came up on Ancestry?

Nicole (2m 57s):
In this particular experiment, I did leave them as they were And I had to prompt the, the Claude chatbot to only look at the ancestors of the main starting person in the tree view that I gave them. Because when I tried this another time, it started opening the index of all the people in the tree and like listing all of them. And it took forever. And I ran out of like my, I reached my limits in Claude and so I realized I need to tell it, I only want the direct ancestors of the starting person in the tree I’m giving you because that’s the person whose DNA matches. And so that was a key thing to do. And then just giving it, you know, look at the surnames in this tree, I think it actually expanded some of the lines itself because it is able to, so it, because it’s an agent, it can click on things in the browser and expand the tree itself.

Diana (3m 51s):
Aha, that’s what I was wondering. Very interesting. Okay, well I’m gonna have to play with that one. I haven’t done any of the agent tools yet, so that sounds like a, a really good use case that’s safe as well. You know, not having it go pay your bills or do anything kind of crazy, but just looking for common surnames. What a great use case.

Nicole (4m 12s):
Yeah, well and the first time you do it, I mean you approve the plan, you adjust the plan and then I just watch it. I wanna really see what it’s doing to understand how it works and, and stop it if it’s doing too much or whatever because it does use up your limits quickly. But I just think it’s so cool that it can do this work for me and I think it is going to be really helpful in the future.

Diana (4m 35s):
So for our listeners who don’t know how to find that, how do you install Claude for Chrome? The agent, how do you do that?

Nicole (4m 42s):
So if you’re using a Google Chrome web browser, you can go to the extension store and search for Claude and then you can see that there. And another way to find it is just to do a Google search for Claude in Chrome extension.

Diana (4m 56s):
And then once you install it, you’ve got the little looks like a spidery orange thing to me.

Nicole (5m 1s):
Yes.

Diana (5m 2s):
Up in your extension bar. So I’ve got it, I’ve got it now I’ve gotta try doing something with it. So I’m glad we talked about this today. It’s going to, it’s going to make me try it again.

Nicole (5m 12s):
Yeah, so then once you click it, it will open a sidebar next to the webpage you’re on and then you can add different, you can drag and drop some of your tabs into the same tab group with that. And anything you put into the tab group with Claude in Chrome, it can read that tab.

Diana (5m 26s):
Okay.

Nicole (5m 27s):
And take actions on that tab for you.

Diana (5m 29s):
Okay.

Nicole (5m 30s):
And it can even, you know, I did test this out too, it can click on links in a research plan document. You know, I gave it a Google doc and told it, click on these links, search these databases for this name and then write a summary of what you found. And it actually did a great job on that.

Diana (5m 47s):
Were those databases, ones that you already had open, like in Ancestry, you know, page that you have your account open or FamilySearch where you already signed in?

Nicole (5m 56s):
Well on my Chrome browser when I go to Ancestry, it’s already signed in, so, so it was able to get in, it was just like I was clicking the link, you know what I mean? So I didn’t click the link for it. I was, you know, just testing out what it could do. But if I hadn’t been signed in on Chrome, it would’ve been stuck, you know, like if I hadn’t already been signed in previously like, like I just keep my Ancestry account open all the time and I’m constantly signed into it.

Diana (6m 21s):
Right, right. Okay, well let’s do some announcements for today for our next Research Like a Pro Webinar Series offering. We’ve got Finding a father for Elizabeth Adcock by Scott Dickson. This will be on Tuesday, April 21st at 11:00 AM Mountain Time. And we are going to learn all about Elizabeth Adcock who was born about 1786 in Granville County, North Carolina. In 1805, she married Henry Vincent and the couple migrated to Rutherford County, Tennessee along with a number of relatives and associates. So far, no direct evidence has been located to definitively identify Elizabeth’s father and family of origin.

Diana (7m 2s):
But ample documentary indirect evidence within the broader family and family network can help to identify her likely father. This case study makes strong use of locality, indirect evidence, negative evidence to support its conclusion. So we’ll be learning all about Granville County, North Carolina; Rutherford County, Tennessee, Wills & Estates, Tax Records, Census, Land Deeds, Indirect Evidence, Negative Evidence. So a little about Scott. He has been seriously pursuing genealogical research since the late 1980s when he inherited several family bibles and many letters including ones from an ancestor couple who went to Africa as missionaries in 1856.

Diana (7m 44s):
His research has focused primarily on the Southeastern U.S., with forays into early Massachusetts and French Canada on his wife’s side. He is a member of the board of the Georgia Genealogical Society and three-time graduate of the RLP study group. Scott works in IT sales and lives in the beautiful Charleston, South Carolina area. We’re excited about the next study group, which begins in August. This will be our Research Like a Pro study group, and the peer group leader application is on our website, join our newsletter. It comes out every Monday to learn all about what’s going on with new posts, upcoming lectures, and any coupon codes.

Diana (8m 23s):
And we’re excited about the National Genealogical Society Conference coming up May 27th to 30th in Fort Wayne, Indiana.

Nicole (8m 34s):
Today we get a fun topic. We’re actually sharing a RootsTech interview that we had with Mark Thompson and Steve Little, all about artificial intelligence, and it was really fun to be at RootsTech and to talk with them. And so they recorded this and graciously shared the files with us so we can also share it on Research Like a Pro. So you might have already heard this interview on their podcast, which is called the Family History AI Show. Well Mark and Steve also write on blogs, so if you haven’t ever seen their websites, you can go to Mark’s blog, which is Making Family History, and Steve’s, which is VibeGenealogy.ai at substack, and Steve’s old blog is also called AI Genealogy Insights, which you can still read as well.

Nicole (9m 16s):
And Mark and Steve also presented in our Research Like a Pro Webinar Series back in 2024. So if you’re wanting to watch that, you can go back and see those webinars if you’re a member of our webinar series from 2024. So they’re wonderful. They’re graduates of our study group Research Like a Pro at DNA and we’re so lucky that we know them. I’m so happy that I was able to learn about AI from Steve. And this is just such a fun interview that we were able to talk about kind of our connection and then just talk about how we’re using AI in Research Like a Pro.

Diana (9m 50s):
Right. It was really fun and some of the things that you’re going to hear about, we are going to talk about how I am a fast follower and Nicole is on the cutting edge of technology. And so it’s kind of fun to discuss that, how Nicole likes to dive into new AI tools. And as you saw in our intro here at the beginning, what we’re talking about that after she figures something out, then I become a fast follower and Mark is actually the one who, who gave us those names and I thought he was just spot on. So it’s fun to think about, you know, how we all have different ways of exploring new things and working with things, especially something like AI.

Nicole (10m 31s):
Another fun topic that we discuss is how mom, you’ve been doing this really cool workflow for locality research, how you use kind of a round robin approach. You gather raw data from all the AI models about a locality and then you feed all that output, which is kind of a lot to distill it down within Claude. So that’s such a fun idea that our listeners will probably get some great practical tips on actually using the chat bots more than one of them.

Diana (10m 57s):
Yeah, yeah, I’ve loved doing that. We also talk about this idea that AI might make us lazy. And so that is a very real fear I know of people. But then we talk about how in the verification step where you fact check AI’s transcription or summary, that’s where you are really going to learn. So you don’t just let AI do all the work you verify and then that is how you learn, how you internalize the information that AI is giving you. So something to really help us to understand how to use it. It just shifts our time from the drudgery of typing to the art of correlation.

Nicole (11m 41s):
Love that.

newspapers.com (11m 43s):
Now let’s have a word from our Sponsor. As flowers bloom and the days get warmer, why not breathe new life into our family history research. Newspapers.com is the perfect companion for Genealogists at any level looking to go beyond names and dates. With access to over a billion pages of historical newspapers, you can uncover fascinating stories about your ancestors that don’t show up in typical records. Picture finding a heartfelt obituary that introduces extended family, a local news article showcase casing your great-grandmother’s community role, or even a photo that captures a special moment from long ago. Little treasures that make their stories blossom before your eyes. Newspapers are packed with the everyday moments, announcements, celebrations, challenges that bring color and depth to the lives of those who came before you. These rich primary sources help you connect with your family’s past in a way that feels personal and meaningful. The search is easy and intuitive whether you’re looking by name, place, or time period, letting you uncover these hidden gems at your own pace, ready to brighten your research with stories that go beyond the usual. Visit newspapers.com/FamilyLocket today and enjoy 20% off a subscription. Let spring be the season you turn your family tree into a lively family

Nicole (12m 46s):
story.

Mark Thompson (12m 49s):
Steve And I are delighted to be joined today by Diana Elder and Nicole Dyer from the show floor at RootsTech 2026. Diana and Nicole are the mother-daughter team behind the world famous Research Like a Pro Books, podcast and genealogy education program. And also I will say two of my and Steve’s genealogy superheroes. So, so nice to have you guys on the show today.

Nicole (13m 13s):
Thank you so much. It’s so fun to be here. And you guys are my AI superheroes. Without you guys, I wouldn’t know anything about AI. I jumped into Steve Little’s class at NGS, Empowering Genealogists with AI and just my eyes were opened and from there I was just amazed and I use it every day now.

Mark Thompson (13m 28s):
Oh, that’s awesome. That’s awesome. So, well like before we get into the AI geekery, you two like, you know, one of the reasons why we really, you know, follow you guys a lot is because you’re a dynamic duo and there’s not too many dynamic duos in the genealogy world. Tell us a little bit about actually how you got together into running FamilyLocket and having this, you know, this, all of these different things that you do in the genealogy world.

Diana (13m 53s):
Well, we were doing genealogy as hobbyists and one day Nicole said, mom, you should get a credential. And I said, really? I can do that. And so I started doing the accreditation process and she kind of followed me along seeing what I was doing. And after two years I wrote Research Like a Pro and then we decided to turn that into a book and then we did a study group for testing purposes and Nicole did the first study group and learned the whole process herself. And then it just has gone from there. We started doing genealogy together in 2003 when she was a teenager. And so we really have learned from together from the very beginning how to do this.

Nicole (14m 32s):
Yeah, it was so fun being a 16-year-old and having this dedicated room up in the, the second floor of our house. And it was, we had moved to Utah, it was cold outside and there wasn’t anything else to do. I didn’t make the volleyball team that year sadly, so I had more time on my hands and it was genealogy that filled in the gaps. So, and it was, was so fun to connect with my mom and learn about her, her dad’s side of the family that I didn’t know much about.

Steve Little (14m 58s):
Mark And I have not been shy, we consider ourselves disciples. There would be no Mark and Steve without Diana and Nicole, and I love hearing about the book and it was the podcast that was the introduction to me. My genealogical education, the foundation was your podcast and I’ve heard every one of your podcasts I would be riding the lawnmower and I would listen to three or four episodes in a row. And then I started to hear about your, you were doing classes. And first I did the asynchronous and I never, never engaged. I, I tried to do it, but it wasn’t until I signed up for the live study group and Mark and I were in the same DNA live study group.

Steve Little (15m 46s):
And that’s where the magic happens. That’s, y’all have the best model for genealogical education. That’s why we steal from the best.

Nicole (15m 58s):
Well we’re so flattered,

Steve Little (16m 0s):
Just so wonderful. And you’ve, how long have you been doing the, the live study groups?

Diana (16m 6s):
The first one was 2017. I got accredited July, 2017 and we tested out the very first one, fall of 2017.

Nicole (16m 14s):
I don’t think we’ve even published the book yet. Had we?

Diana (16m 16s):
No we hadn’t. I was just like, this is amazing, this process I learned in accreditation and I wanted to see if it would help anybody else. And so that first little study group I think only had 12 people and we charged a very minimal fee and we…,

Mark Thompson (16m 30s):
You test your pack

Diana (16m 31s):
And we had beginners, I had my neighbor do it as a payback for quilting for me. And she was a literal beginner. And then we had some other people who were actually pretty good researchers and for every single one of ’em they finished up with a report with citations. It was so gratifying. And then we’re like, okay, this works. So cool.

Steve Little (16m 50s):
I think what the magic ingredient was for me, and I’m sure I’m not unique in noticing this, the community, you build community. You’re not just sharing information but you, the cohorts, I’m still in contact, well with Mark, but all the other people from my cohort. And so you get that a little bit on a podcast. You probably have people come up to you and you don’t know them, but they feel like they know you very, very well. I do that when I see George or Drew Smith across the hall somewhere. He doesn’t know me from Adam, but I know his voice in my ear from the lawnmower And it, and so we’re with the whole community is very grateful for the communities that y’all are building.

Steve Little (17m 35s):
So very thankful for that.

Mark Thompson (17m 38s):
So the the, the Research Like a Pro methodology, I mean I consider it the methodology for doing genealogy. You know, it’s because it got burned into my brain so many times after taking the class several times ’cause I liked it so much. But when did you first realize that AI could actually be helpful? Like there was a, there, there for whatever it was that you did, whether it was personal or professional.

Diana (17m 59s):
Well I have to say it was Nicole who discovered that first because she is more the experimenter. Mark and I have talked about this.

Mark Thompson (18m 6s):
Yeah.

Diana (18m 6s):
And he gave me the ranking of fast follower, which I absolutely agree with. And Nicole is cutting edge. So I followed her.

Nicole (18m 16s):
It’s funny because I was just coming out of having my fifth child and it was a difficult birth and, and I was actually in the hospital I think when I heard about AI and Steve Little and doing things And I was like, that’s so cool. And I was at a conference where Angela McGee had said, we’re doing this course at NGS and and I’m gonna do it. And I was like, oh, I wanna do that. I signed up right that day when she told me. I was like, yeah, that sounds fun. And I hadn’t done any classes in a while ’cause I’d just been focused on family. And so I was like, yeah. And the course was just amazing. The way Steve taught about the four AI transformations resonated with me. So well just like, these are the things AI can do. Oh, like I can apply this to so many things that I do regularly as a professional genealogist and educator.

Nicole (18m 60s):
And it just clicked.

Diana (19m 1s):
So what I like to do is try the same prompt and I think you guys teach this, try the same task in all the major large language models to see which one worked best. And I did that when I was learning so that I could kind of just see which one I want to use for which task.

Mark Thompson (19m 17s):
Oh, of course. Yeah.

Diana (19m 17s):
And that’s been very helpful.

Mark Thompson (19m 18s):
So, so did, which one did you settle on was your happy place for locality research? For your locality prompt?

Diana (19m 24s):
Locality prompt. So I tried Deep Research in ChatGPT, gosh chat sheet BT And I got a 250 page one that was way too big. Like I don’t need 250 pages. So I tried Perplexity. I actually really like Perplexity’s deep research ’cause it seems to keep it not so long, but pretty, pretty good. So you know what I do now you’ll laugh at this, but I do it with all of them. I do all four with the same template. I take all of those and I put those into Claude and I say, okay, just make this nice for me. Just gimme the most salient points in the links.

Mark Thompson (19m 56s):
I am, I’m not laughing at all. ’cause I do that with a lot of the things I do is like, you know, to do the consolidation. Yes. Consolidate the output across a bunch of them. ’cause they’ll give you such different results.

Diana (20m 5s):
I know. Yeah. And then I wanted Claude as my favorite for formatting and writing. Yeah. And so, and then I wanted Claude to just fix it all up. So it’s how I wanted it.

Steve Little (20m 13s):
It’s so interesting to hear that we are, I, I’m realizing, probably not for the first time, that we’re independently discovering workflows that really work. And, and what I heard y’all describing was at initially perhaps gathering too much information or information from various sources and then having one of the chatbots consolidate that information, integrate and synthesize them and then distill it down to one 10th of its original size. And that’s magic. And and then it’ll have linked sources to all the original sources. And like Mark taught yesterday in his research workshop, they’re really good.

Steve Little (20m 56s):
Those you have to double check all the links but it’s rare when, when it’s not correct.

Diana (21m 1s):
We’ve gone from just going to Google and saying, okay, tell me everything about this county to having AI go find it all for you. It’s so much better.

Mark Thompson (21m 10s):
Yeah, yeah. And faster and more complete.

Diana (21m 12s):
Yeah.

Mark Thompson (21m 13s):
Right. You still gotta verify the heck out of it.

Diana (21m 15s):
You do and you know, I am always questioning, okay, am I learning as much?

Mark Thompson (21m 19s):
Yeah.

Nicole (21m 19s):
About the county?

Diana (21m 21s):
Yes.

Nicole (21m 22s):
Or whatever.

Diana (21m 22s):
And so I think there’s that to balance. You know, how much we are understanding the information versus how much AI is just spitting it out and we’re just putting it in a file.

Nicole (21m 31s):
I think that’s where the fact checking comes in. That’s where we really have to learn it because we’re checking it. And if we skip that step, we miss valuable learning and we introduce hallucination.

Steve Little (21m 39s):
Oh that’s a great insight. Yeah. That it’s that if you were concerned about that cognitive evaporation, not learning from doing the research yourself, that it’s during the verification step. If you are concerned about, you know, learning the material, the verification step is when you can actually make sure you’re absorbing it yourself. Yeah, that’s great insight.

Nicole (22m 5s):
That really serves two purposes, doesn’t it?

Steve Little (22m 6s):
That’s a great insight.

Mark Thompson (22m 7s):
I, I see people go through this, this arc of, ooh, I can get more words now than I used to be able to get myself. And they end up with a whole different problem. Right. Like all of the like like with FamilySearch, the new FamilySearch Full Text Search, I went from, I searched for hundreds of hours to find five or six probate records for my people on this one line that I was trying to find the answer for. And then when Full Text Search came out, suddenly I had a data overflow problem ’cause I, instead of finding five, I found about 35 one day. And now I went from like trying to glean every single fact to, I needed to build a workflow to make sure I didn’t lose it and that I could keep track of it and what I was gonna do follow ups on.

Mark Thompson (22m 50s):
I ended up with a a completely different workflow of problem than I had two weeks earlier. To, to that end though. How have you found that since you’ve started to introduce AI into the Research Like a Pro process, how has it actually changed what you do? Like you still follow the process and you’ve AI enabled it, but has it shifted where you put your time and your efforts differently because you’ve introduced AI?

Nicole (23m 13s):
Oh yeah, because I used to spend a lot of time transcribing records manually and now I use Gemini. Yeah. And it’s so much faster. I still fact check everything, especially names and dates and places. Yeah. But I can spend more time writing which, and correlating, which is the part I love. Putting evidence together and seeing the answers come to light, planning other things I need to get to find original sources, you know, beyond the ones that are online. I ordered some records from the National Archives the other day. They had the direct evidence I needed to get a parent-child link. That never happens for me. I was so excited. I had spent hours transcribing the land patents and nothing, you know, tax records and deeds and nothing had connected the men together as parent child.

Nicole (23m 58s):
They were just living right next to each other for a long time. But in the National Archives it had that direct evidence in one of the witness statements in a cash entry land patent. And so I love that part, being able to, instead of focusing so much on transcribing digital sources, AI can help me get that done faster and I can then order these other sources and I have the bandwidth to be able to work on getting those from the repositories where they’re not digitized.

Steve Little (24m 26s):
I love that, you it, it makes some parts more efficient so that you’re able to spend more time on things you wouldn’t have been able to have time to do before.

Nicole (24m 36s):
A hundred percent

Steve Little (24m 36s):
Or just would’ve taken much longer to get to.

Diana (24m 38s):
One of my favorite use cases is court records because I came across a really fun assault and battery case. I shouldn’t say that’s fun, but

Nicole (24m 49s):
It was entertaining.

Diana (24m 50s):
It was interesting and entertaining. And so I had used AI to, you know, do what we do transcribe and, but it still was not making sense. There were a lot of different dates, you know, like this money was owed and this sheriff had to go out on this day and then they came to court and we had this defendant and that defendant. And so I put it all into AI after I had it completely transcribed, you know, these very cryptic minutes with a lot of shorthand by the clerk, and AI chronologically put that together and then told me what happened. And of course then I could go and verify and say I agree or I don’t agree, but it made sense. And otherwise that would’ve taken me a long time to try to understand. And there were some really interesting terms.

Diana (25m 30s):
I didn’t know what those meant and AI knew exactly what those meant. So court records, it has just opened those up like nobody’s business

Mark Thompson (25m 39s):
Analyzing technical records, court records, military records, medical records, like legal records, like the, it’s just completely revolutionized being able to, ’cause you can, it’s all in context now. You don’t have to learn the lingo.

Diana (25m 50s):
Yeah.

Mark Thompson (25m 50s):
Right. You can get it to bring its lingo and, but then, then you of course gotta check it. But man it’s just fabulous for that.

Nicole (25m 57s):
It gets you started,

Mark Thompson (25m 59s):
Gets you started. So one of the other areas of course that is really challenging for a lot of people in the genealogy world is DNA and DNA evidence and, and people spend years sometimes just getting the real fundamentals of DNA before they can get back to doing the really good genealogy research. How has AI or has AI changed how you do the DNA parts of the Research Like a Pro process?

Nicole (26m 25s):
Well, we’re still learning and it’s a slow process because we wanna be very careful about privacy and data training and how much do I trust the AI tools with these sensitive names and amounts of DNA. So I’ve been really trying to understand that and doing a deep dive into what do I need to do to implement AI with DNA before I can feel comfortable. And when Claude changed to training on your data by default, I didn’t really understand that that was happening. I think probably just didn’t read the whole terms of service when it popped up. So there was a few months there where I thought it wasn’t training on my data. So I think we just always have to be aware that things change in AI to, before we do anything sensitive with DNA check and see if data training’s on, go to the settings, find that setting, make sure you know what it’s doing.

Nicole (27m 13s):
And, and also just there’s a lot of ways to use AI for DNA without putting anything about living individuals in the AI tools. I like to anonymize names pretty much all the time before I put them in. Or just ask a question that doesn’t have a name attached. Like can you analyze these, What Are The Odds scores and just help me understand what it means that this one is five times more likely, but the other one is 400 times more likely? And just help me talk me through that. Look at the, the frequently asked questions on the website about WATO and, and that kind of help is really good right now for DNA but the cutting edge things are that it can make a network graph. I mean that was really cool. And it can help with pedigree triangulation.

Nicole (27m 54s):
I mean I’ve tried Claude in Chrome as a sidebar adding three public member trees to the tab group and saying find the common ancestor.

Mark Thompson (28m 1s):
Yeah.

Nicole (28m 1s):
Like that. And it did it,

Steve Little (28m 3s):
And I appreciate your honesty, that Claude caught us off guard when they changed their terms of service and started using user data for training. Right now we have to upload a lot of data to the AI to get it to process. Not too long, this concern will evaporate. When we do this on our local computer and none of it leaves our local computer we can breathe a lot easier, but we’re not there yet and probably won’t be for a year or two at least. But in the meantime we can, there’s, if there’s a will, there’s a way as my mom would say. And so if you know what the dangers are, these tools can actually help you avoid the thin ice and find a safe path.

Nicole (28m 50s):
What a great idea. And it just made me think about uploading data to Claude and ChatGBT. But there’s so many tools now with the AI baked in with enterprise level security and privacy. Goldie May, Airtable, you know, we love Airtable. And they have a whole sidebar chat bot now Omni, which is very good. And if I’m tracking DNA matches data there and I ask Omni to correlate, I haven’t tried this, I need to. But it is amazing and the field agents in Airtable can analyze an attachment, transcribe it for you with Gemini and the field attachment, like they, the field agents can do so many things with your data in Airtable. Oh that’s awesome. And then it’s secure and private. I love it.

Mark Thompson (29m 29s):
Have you guys tried out to the new Goldie May AI yet? I’ve been meeting to go by the booth and get the, ’cause I haven’t been a Goldie May, I’ve, I experimented it with a couple of years ago and I haven’t gone back to it since they, they just recently announced their AI and an AI enhanced version. What’s your, what’s your initial thoughts?

Diana (29m 44s):
Yeah, it’s really fun. You definitely need to go try it out. And the developer, Richard Miller, has an hour long video that you can just watch him in action and he says, this is not edited at all. This is just me in action. And it’s fun to watch to see what he’s developed and to try it out. So yeah, that’s another good example where you just go use his tool and it’s going to work with,

Nicole (30m 6s):
well it’s connected to the FamilySearch tree, and so you can actually put in two ID numbers for people who like maybe you built one of your DNA match’s tree to a deceased person who died in 1940 and your grandpa who’s deceased, ’cause the family search tree works on deceased people. But in there two ID numbers into Goldie May AI assistant and say, search the FamilySearch tree and see if these people have a common ancestor. It was literally two seconds later that it found the common ancestor. And of course I chose a test where I already knew who it was, the common ancestor was, but it worked. And I’m like, this has a huge potential.

Mark Thompson (30m 40s):
Like this is what 2026 and 2027 is gonna look like for me. Like companies that build products that would benefit from AI enhancement that understand how a Genealogist’s brain works. Like Goldie May, they’ve been doing gene supporting genealogy tasks for years, so they really understand what if we could add a little bit of AI special sauce in that spot, that would be just magical. So if, when you look forward, when you guys think about the boy, I wish they could add some AI, you know, pixie dust in this place, in that tool or at that AI company. What have you thought about? What, what is the place where you really wish AI could help?

Diana (31m 21s):
Oh, well I would like to have Ancestry, have AI in there with DNA and say, ask a question about my DNA matches and just spit out the answers.

Mark Thompson (31m 29s):
Yeah, wouldn’t that be nice.

Nicole (31m 30s):
Everyone does. I saw someone in the Facebook group asking how can AI find my biological father? And I think down the road that might be a thing, but right now it’s gonna take someone who knows how to analyze DNA evidence and documentary evidence to guide the AI tool to get the result you want. But down the road I can see that, you know, futuristic application of that.

Mark Thompson (31m 50s):
I, I think there’s so much possibility in the DNA world, well particularly with now MyHeritage and FamilyTreeDNA, they’re going to next generation sequencing. So the amount of information that’s two, for two reasons, more information than we’ve ever seen before by about three or four orders of magnitude. Like that’s gonna just blow people’s minds to start with. But now, instead of having just three types of DNA to worry about four, if you count X, right? Now, we’ve got three different types of autosomal DNA too, and people are gonna be like, what is, oh, Ancestry does it this way, FamilyTree does it that way, myHeritage does it that way, what does this all mean? Like if they had a tell me what my results mean button like, and it would just give a summary, the type of thing that a, a genetic genealogist would do a quick overview in an hour with somebody, that kind of stuff I think is implementable with a single prompt button.

Mark Thompson (32m 44s):
The general response that, and if they did it, there’s no security problem. If we do it, then we’ve got a security problem, ’cause we’re sharing data, right? So I think that is like if I, if I could wave my magic a AI wand, boom, gimme the explain my DNA results button.

Nicole (32m 58s):
I bet that’s coming.

Steve Little (32m 60s):
Mine would be, see if I had a, a magic button like that, it would be for audio and video. Working with the audio and video yet is not quite as easy as it ought to be. Everybody has audio transcripts or VHS from 1950 or 1970s or eighties, you know, the family reunion or the wedding or the Christmas when grandma was talking about her grandma. And, and it’s not easy now to get that into a genealogical database. Not as easy as drop and drag, but we’re almost there. So just being able to drag a MP3 or MP4 file onto a, a button that says, what does it mean and turn a a, a audio file into a family tree and a story and a genealogical proof statement.

Steve Little (33m 52s):
I can see that just over the horizon.

Mark Thompson (33m 54s):
I was wondering, so we’ve talked about some amazing things that AI can do and some new areas that we wish AI could help us in, what do you think are the types of things that AI is never gonna replace? Like what are the parts of like the secret sauce that make Genealogists love to do genealogy that AI will never get to or that we hope it never gets to?

Diana (34m 13s):
I hope it’s not going to replace us and our brains and our reasoning, our conclusions, you know that where would the fun be?

Mark Thompson (34m 21s):
Yeah.

Diana (34m 22s):
Because I love taking the data and making it mean something. I like AI to help me, but then I wanna be the final say. And of course we can just still do that. But I would hope that that wouldn’t replace humans.

Nicole (34m 35s):
I really don’t think it will. I just think genealogy is too nuanced. And I have friends that ask, is AI gonna replace me as being a professional genealogist? Will people be able to do this advanced work on their own now with all the advancements? And I just don’t think that it will. I think that it’s going to help us become more efficient, but I don’t think it’s gonna replace us. And I can’t put my finger on why, but in all my testing with AI, when it comes up with research plans, when it follows research plans, it is never as good as I am.

Mark Thompson (35m 5s):
Yeah.

Nicole (35m 5s):
So I just, I can see that it’s getting better and it’s probably gonna get better, but there’s just a human element and maybe it’s a bit of just inspiration that we have that the computer doesn’t have. We get these aha moments, these ideas, maybe we should try that. We are more determined. We have unlimited context windows to keep working all night long and we remember everything from the beginning of that day. Three months later we forget it. But I just think there’s that human element of persistence, determination, and inspiration that the computer will never replace.

Mark Thompson (35m 40s):
Well that’s wonderful. It has been so nice to get together with you guys in person again this year at RootsTech. Thank you very much for sitting down with us to talk today and we can’t wait to do it again next year.

Steve Little (35m 49s):
Thank you. Thank you. Thank you.

Nicole (35m 51s):
Yeah, let’s do it again next year for sure. Thank you guys so much.

Diana (35m 57s):
Thank you. Okay, well that was so fun to hear that interview with Mark and Steve. We loved being in expo hall with them and hearing all the fun noise in the background brings that back.

Nicole (36m 6s):
One of the things that I loved about recording this interview with Mark and Steve was just how easy they were to talk to about AI and just, Steve is so good at taking an idea that we expressed and then summarizing it and kind of pulling out the principles that are relevant to us in genealogy. So it’s so fun to see Steve do that and I highly recommend that anyone listening watch Steve’s RootsTech class. It was amazing and there’s so much that you can gain from learning about AI from a true expert. So make sure you watch that recording. And Mark’s classes are also amazing and you should watch his too.

Nicole (36m 48s):
We’ll put links to them in the show notes and hopefully you guys can take a minute or two and watch their classes. They also have courses that you can join by purchasing and they have a beginner’s class and an advanced class. They have their advanced class is coming up and I’ll be taking that because I just find so much value in learning from both of them. So I highly recommend their courses. So I hope you guys will try them out too.

Diana (37m 20s):
So we hope you get something out of this interview about how we all use AI and our thoughts on that. And if you haven’t started to use it yet, we hope it’ll encourage you to try a little bit. So thanks everybody for listening and we’ll talk to you next time.

Nicole (37m 35s):
Alright, bye-bye. Thank you for listening. We hope that something you heard today will help you make progress in your research. If you want to learn more, purchase our books, Research Like a Pro and Research Like a Pro with DNA on Amazon.com and other booksellers. You can also register for our online courses or study groups of the same names. Learn more at FamilyLocket.com/services. To share your progress and ask questions, join our private Facebook group by sending us your book receipt or joining our courses to get updates in your email inbox each Monday, subscribe to our newsletter at FamilyLocket.com/newsletter. Please subscribe, rate and review our podcast. We read each review and are so thankful for them. We hope you’ll start now to Research Like a Pro.