In this episode of the Research Like a Pro Genealogy podcast, Diana and Nicole discuss using Airtable AI tools to evaluate DNA matches. They explain that Airtable AI can make writing about DNA much more efficient and explore how these cutting-edge features can be integrated into genealogical workflows.
The hosts focus on Airtable AI’s capabilities in long text fields, where users can write paragraphs, summarize information, and extract details into lists. They provide step-by-step instructions on how to set up and use these AI tools, including creating custom prompts.
Diana and Nicole also discuss the different Airtable plans and their associated costs and credits for AI usage.
Listeners will learn how to use Airtable AI to analyze DNA match details, such as shared DNA and hypothesized relationships. The hosts provide a practical example of how to use AI to analyze a DNA match between Nicole’s grandfather and his brother. They highlight the advantages of using Airtable AI, including its seamless integration with existing data and the reduced risk of AI hallucinations. Finally, Diana and Nicole explain how to use the Writage Word plugin to integrate AI-generated analysis into a research report.
This summary was generated by Google Gemini.
Transcript
Nicole (1s):
This is Research Like a Pro episode 321 Using Airtable AI Tools to Evaluate DNA Matches. 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 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 (41s):
Today’s episode is sponsored by Newspapers.com. Hi to all of our Research Like a Pro fans, welcome to the podcast today.
Diana (49s):
Hi Nicole, how are you doing?
Nicole (52s):
Good. I am just happy to be talking about Airtable today. It’ll be fun.
Diana (56s):
It will be fun. And I’ve been exploring it as I’ve been working this summer on my artificial intelligence Research Like a Pro project and learning how to use it. So it will be fun to talk about this. Our announcements for today, we have our September Research Like a Pro webinar coming up on the 21st at 11:00 AM Mountain Time. That’s a Saturday. And this will be by Karen Ramon, “Proving the Parents of John G. Winn: A 19th Century New England Study”. An undocumented history gave clues to John G. Winn’s parents. Using the Research Like a Pro method, those clues turned into evidence!
Diana (1m 37s):
The development of a strong locality guide and a journey through courthouses and cemeteries, combined with online sources and a town clerk’s office, led to various records that helped conclusively document John’s connection to his parents. And the topics are Massachusetts, vital records, census, newspapers, cemetery, probate. And I am excited about this because in New England we often see these genealogies that give all the lineages, but not a shred of evidence, no documentation. And so it’ll be really fun to see how she backed that up. Well, we’re excited to announce that our next Research Like a Pro DNA study group will be beginning February, 2025.
Diana (2m 22s):
And if you’re thinking about joining us, you can check out the Peer Group Leader application on our website, see if you’re ready to apply for that. And registration will open in December. We invite you to join our newsletter so that you can get the latest in the blog posts and videos and podcasts, as well as information about where we will be speaking and other fun things. So fall conferences that are coming up. We have the APG Professional Management Conference on September 19th through 21st, which is Virtual. Nicole will be presenting there. And then the East Coast Genetic Genealogy Conference, October 4th through 6th in Maryland and also online.
Diana (3m 3s):
And we’ll both be providing some lectures there. And then I will be doing three presentations for the Texas State Genealogical Society Family History Conference, November 1st through 2nd. So one each month coming up in the fall, which will be very fun. We hope to see you at some of those.
Nicole (3m 23s):
Yeah, it’ll be nice to have one each month, like you said. Well, in the ever evolving field of genetic genealogy, we’re always looking for innovative ways to analyze and interpret our DNA data. And as you know, Airtable.com is one of my favorite tools for this. It’s a very versatile database platform. It’s really flexible for logging DNA evidence in genealogy research projects. Earlier this year, Airtable announced they have AI tools integrated within Airtable. So that was really exciting and we were super excited in July when we heard that Airtable AI tools would be free for all users. But we’ve since kind of learned that only some of the AI tools are on the free plan and the one that we really want the generate text.
Nicole (4m 8s):
one is not on the free plan. So you do have to have a paid plan sadly, but after you learn about what it can do, you might be willing to pay for that. Well, what it can do is it can help us make writing about DNA much more efficient. And so in this episode we’ll explore how these cutting edge AI features can be integrated into our genealogical workflows, especially within our Research Log in Airtable. And when I wrote this blog post that we’re using as the text for this episode, I had talked about Airtable AI in a lecture. And so I used some of my notes from that lecture and asked Claude AI to help generate this blog post. So just to shout out to Claude for helping me put this all together and into words and sentences.
Nicole (4m 52s):
Sometimes it’s nice to be able to take something you’ve already done and put it into another format using AI.
Diana (4m 58s):
I think that’s such a great use of AI because you’ve already created the content, it’s all your ideas, it’s just getting it written in a little bit different way. Well, let’s talk about Airtable AI and the game changer for long text fields. So one of the fields that we do love using in our Airtable research logs is called Long Text. And it lets us write as much as we want in that little cell. And you can open it up and paste in an entire paragraph from a county history or more. You can type as much as you want. So according to Airtable’s support documentation, these tools allow users to write paragraphs, summarize information and extract details into lists.
Diana (5m 46s):
So to harness these features, we can add or reconfigure a long text field in our Airtable base, and then we can toggle on the generate text button in the field settings. Then we can create a custom prompt preview the AI generated output, and then choose between automatic generation or manual activation. So as in all AI, we have to figure out our prompt, and our prompt will guide the AI in doing what we want it to do. So in Airtable to create a prompt that includes data from various fields in the row, you type the curly bracket symbol and it will show a list of all the fields in the table.
Diana (6m 34s):
For instance, in our Research Log locality, date, the person, the citation, you know, all those different things that we fill out. And then after you select the field to draw the data from, the selected field will have an oval around it so that you know you’ve got it selected,
Nicole (6m 55s):
Right? So it’s just a pretty easy way to use your Research Log and tell a large language model to, Hey, take these parts of my row and create a paragraph. And we kind of give it some instructions as if we were telling an assistant or an intern, Hey, I want you to write a paragraph from this Research Log, here’s how you would do it. And you’re gonna start by saying, 1850 census. And this person in the people column or field was listed on the census, and here’s the citation to add as a footnote. So you can tell it all those things in your prompt and then it will do that. Well, I mentioned that we were thinking Airtable plans were going to allow free users to use these AI tools to generate text and to write sections of your research report for you.
Nicole (7m 42s):
But in our workshop we found out that those 500 monthly credits for the free users didn’t include the generate text option in the long text fields, which is the one we wanted to use. So in order to use this feature, you need to be on a paid plan at Airtable. Those paid plans start at $20 a month for the teams plan. With that, you get 500 monthly credits to use and those credits do not roll over each month. So if you’re not using it, you might want to cancel your monthly plan until you’re ready to use one and then sign up for one month. If you need to use more than 500 credits, then you can do the $6 add-on for more credits.
Nicole (8m 22s):
And the way that the credits work is that if you have a long prompt and a long response, it uses like 20 credits. If it’s more of a shorter prompt and a shorter response, it’ll use about 10 credits and it will show you in the text edit preview field, when you type in your prompt and shows you a preview, it will show you how many credits it thinks will be used for that prompt and response.
Diana (8m 44s):
And it would be really wise to look at that because I used up all of my credits in one, you know, generating some text, and I had put in there a really long part of a county history and all of a sudden it used to all my credits because I had too much text. So you know, I was just the beginner learning to work with it.
Nicole (9m 3s):
Oh, did you do the automatic generation where you just automatically generate all of the the rows in your log at once?
Diana (9m 10s):
Well, it did one at a time. It wasn’t automatic, it was one at a time, but I hadn’t opened up to see how much text was actually in, you know, my results box. And I had put in a bunch of county history stuff that was kind of long and I’d already generated a lot of other rows, so I just, I just thought I, it would just go on forever and over and let me keep using it. But new I learned my lesson.
Nicole (9m 34s):
Yeah, so I would always select manual activation, and that means you’d have to click the button generate in each row when you’re ready to generate the text for that, because then it allows you to make sure you have all the fields in the row correct or up to date or fully filled out before it generates the text and uses your credits.
Diana (9m 52s):
Very good tip. Well, let’s look at a practical application and let’s look at analyzing DNA match details. So we have a lot of data in our Airtable Research Logs when we’re working with genetic genealogy. So using the Research Like a Pro DNA 4.0 Airtable base, we can add an AI field and automate the analysis of shared DNA and hypothesized relationships between DNA matches. So this is so great. We can set up a prompt so that the AI will analyze the range and average from the Shared cM Project, calculate standard deviations, apply custom evaluation criteria and consider possible reasons for outliers.
Diana (10m 37s):
So this is just awesome. For instance, we might instruct the AI to note that if a match is below one standard deviation from the mean, it could indicate a half relationship instead of a full one.
Nicole (10m 48s):
Great. So I did put that in my prompt that for the last section of the paragraph that I’m having AI generate that if you know it’s something significantly below one standard deviation from the mean, then it could be because of the half relationship option. And then if it’s too high, it could be because of multiple relationships shared with the two test takers. So just giving it some things that it can note when it’s doing the analysis for why it might be higher. To help us think about why there’s outliers and what to do about it. Our Sponsor today is Newspapers.com, your go-to resource for unlocking the stories of your ancestors. Dive into the newspapers where your family’s history unfolds as you search nearly a billion pages in seconds.
Nicole (11m 34s):
Newspapers.com offers an unparalleled treasure trove of historical newspapers providing a window into the past with papers from the 17th century to today. Newspapers.com is the largest online newspaper archive. It’s a gold mine for anyone seeking to uncover stories from the past. Whether you’re a seasoned genealogist or just starting your journey, Newspapers.com makes it easy to search for obituaries, birth announcements and the everyday stories that shaped your family. It’s like having a time machine at your fingertips. And here’s the best part, our listeners get an exclusive offer. Use promo code FamilyLocket for a 20% discount on your subscription. That’s FamilyLocket at Newspapers.com. Sign up today at Newspapers.com and embark on a journey of discovery.
Nicole (12m 15s):
Well, for an example of using Airtable AI tools, I selected a match between my grandfather and his brother. And so I would expect them to share the amount of DNA for full siblings. So I had entered in how much DNA they share in the row and the relationship of siblings. And then in my AirTable base template, Research Like a Pro with DNA 4.0 it brings in the Shared cM Project data from one of the tables within the base. And so it can bring in all of the range for that sibling relationship, the standard deviation, the mean, and all of that. So then when I click generate in the AI field, it uses the prompt that I’ve put in about analyzing all this information and telling me reasons why it might be high or low in that AI field.
Nicole (13m 1s):
I click the generate button, which I named that field to be AI analysis of match. And after clicking on generate the AI tool used some of my allotment of credits to generate a response. And one thing to note is that you can only click generate when all the fields used in the prompt are filled out for the row. And recently I kind of figured out a workaround for that because sometimes they don’t always want to fill out every single row. So sometimes I’ll just put NA into a field to get Airtable AI to still generate the paragraph when I don’t wanna fill that that field out. Well, after the text generates, you can expand the cell to read the entire analysis.
Diana (13m 43s):
Well, I had to do the same thing with putting, I think I had to put none, because it was the fans club column or field and some of my records, I didn’t have any fans and so I had to put none in those so that it would not work. So interesting how we can figure out those workarounds. Well, let’s talk about some of the advantages. If this sound is sounding kind of hard and you’re wondering why in the world you would wanna do this, one of the most compelling advantages of using Airtable AI is that seamless integration with your existing data. So you are generating texts within the same environment where your data resides. You are the one entering in your data into your Research Log.
Diana (14m 24s):
And so you’re going to significantly reduce the risk of AI hallucinations where AI might produce something inaccurate because they don’t have the context. You are giving it the exact context. So this makes your analysis more accurate and it also simplifies your workflow so you don’t have to transfer your data to your research report by little pieces. You can have it all generated and then you can put it into your report in one fell swoop. So it’s a lovely thing.
Nicole (14m 58s):
Great. And I think I will go ahead and read to you a bit of the prompt and a bit of the response just so you can see how well it works. So the prompt that I wrote in to the generate text, long text field and Airtable is discuss whether this DNA match falls within the range of one standard deviation from the mean using the Shared cM Project, state the names of the, and then this was one of the fields, test taker and, another field, DNA match state, the amount of centiMorgans and relationship. And then I have a bunch of fields that I, I’m telling it to pull in here and the range of expected centiMorganss as well as the average centiMorgans, which is the same as the mean and the standard deviation.
Nicole (15m 44s):
Next, review the information from the within range column and within one standard deviation column for that relationship. If the amount of centiMorgans from the centiMorgans column is outside of the range, the relationship is probably incorrect or there are multiple relationships. Next, review the information from within one standard deviation column for that relationship. So continuing on, I just kind of give it some instructions about how exactly to write the paragraph. So then once it reads all the information in the row and follows the prompt, the output is a section of a research report that you can copy and paste into Word or into a Google doc. And so expanding the cell for my AI analysis of match column, it now says for my grandpa and his brother to determine whether the DNA match between Charles William Elder and Richard Elder falls within the range of one standard deviation from the mean using the Shared cM Project we need to analyze the provided data.
Nicole (16m 42s):
(1) Charles William Elder and Richard Elder are Siblings. (2) shared centiMorgans amount, 2527, (3) expected range for siblings. Then it has the range, the mean, and the standard deviation. So it has all this data. And one thing that I have noticed about these language models is that they love to do lists and bulleted lists. So that’s what it kind of de defaulted to is a list of data. And then towards the end of the list it has some analysis. It says the shared centiMorgans amount of 2,527 centiMorgans falls within the range of 1613 to 3488 centiMorgans, which is the range expected for siblings.
Nicole (17m 23s):
It also falls within one standard deviation. So it does the analysis for you. The great thing about this is that it allows you to think about how you would write the paragraph once when you write the prompt and then you can generate a similar paragraph or section multiple times using similar data from all the rows in your log. So it really automates the process, but you have to put in the effort to think, how would I want this to be written? And then once you’ve thought about that and written it out, the rest is easy.
Diana (17m 54s):
I love this application of DNA because we have so much data in our Airtable DNA Research Log in the match details page. And if we can get AI to just bring that all together in a nice way, that will save so much time in our writing. So I was wondering if you could take those bullet points, and I know you can do this, I’m just wondering your thoughts on it. What if you take that bulleted list, put it into another large language model like Claude and say, can you write this in a paragraph, you know, a narrative so that something besides bullet points to put in your report?
Nicole (18m 29s):
Yeah, absolutely. But you know, I probably would just adjust my prompt in Airtable to say that instead of not mentioning the type of output, and as you know Steve Little’s idea for prompting where we have like a role, a goal, a task, and a flask, I didn’t really provide the flask very clearly. So the flask is just the output that you want it to be in. So I would just maybe add a sentence to my prompt that says, please write this in sentences and paragraphs in a narrative style without bulleted lists.
Diana (19m 1s):
Exactly. Perfect. Well, it’s really fun to see what you can do. And I love this example, and I love the idea of being able to use my Airtable Research Log to write reports because I’m in control of everything I put in that. And being able to just easily have the AI assistant put that into a narrative or a bulleted list for me is wonderful,
Nicole (19m 24s):
Right? It’s kinda like writing a template that then someone follows for you. So you’ve basically written it, you’ve supplied all the data and it’s just doing some of the grunt work for you.
Diana (19m 37s):
Our assistant.
Nicole (19m 37s):
Right? So once it’s generated for you, then the next step would be to take that generated text out of Airtable, copy and paste it into the program you’re using for your report, which usually for us would be Microsoft Word or a Google Doc. And we’ve found a couple tools to make this a little bit easier. So the way that the text generates from a language model like ChatGPT or within Airtable, which also uses the GPT models and the Claude models, the way that you can see the text generated is often in a markdown format, which is a simple markup language for indicating how the text should be formatted. So if something is a header or a bold or italic, there’s certain little symbols.
Nicole (20m 18s):
And so you can see in the example, in the blog post that there are two asterisk surrounding names and relationships, and that means that that should be in bold. So all those little things in markdown, when you copy and paste that into Word, you don’t want to see the asterisk, you want it to automatically be formatted the way that it formatted in the language model. So the way to do that is to use a tool called Writage, W-R-I-T-A-G-E, Writage. And this is a tool that you can try for 14 days for free and then purchase for a one-time fee of $29. And once you do that and install it into your Word program, it’s on the ribbon and the toolbar in word so that you can automatically paste markdown, and the formatting is converted into standard Word document formatting.
Nicole (21m 9s):
So I highly recommend this tool. I’ve been using it for a while and I love using word for my research reports. So it makes it really easy to copy and paste from an a large language model into Word, and it looks professionally formatted right away, so you’re not having to do any extra work. And it can even do your footnotes. So if you put in your prompt that you would like to have a footnote with a source citation, then when you paste the mark down into Word, it will automatically put it in a footnote. And then each subsequent paragraph with footnote that you paste in, it will number those footnotes automatically. So it really saves a lot of time and and typing. And then I also discovered recently that there are some add-ons for Google Docs that you can add.
Nicole (21m 52s):
And when you have those implemented in Google Docs, you have a new menu item on your right click menu that says Paste with Markdown, and then you can also paste right into Google Docs and preserve that formatting.
Diana (22m 5s):
Isn’t it great, all these tools that are just coming up for us, you know, we didn’t even know anything about Writage, or maybe you did. I didn’t know anything about it.
Nicole (22m 13s):
Well, I never needed it until I was using a large language model to generate some text,
Diana (22m 19s):
Right. And I did use this for writing my report and it was a beautiful thing. It saved me so much time. It was so great. Well, let’s wrap this episode up a little bit with just doing some final points here. So I think everyone listening can see that integrating Airtable AI tools into our DNA genealogy research is huge. This is so great for us. It’s taking complex analysis and streamlining our documentation. These tools really will help us to focus more on interpreting the results and making those connections in our family history that we are trying to do with our genetic genealogy.
Diana (23m 0s):
So both for people who are experts at this or have been doing this a while and even, you know, those who are just getting started, this really gives some opportunity to help us with our research and uncover some hidden connections. You know, and Nicole, as you were talking about the standard deviation and having it cue you that maybe this is too much, maybe there’s another relationship or it’s not enough DNA shared, perhaps it’s a half relationship, that’s such a great thing to have it remind you of because sometimes we just forget to put that in our reports. So it’s so nice to have it, you know, in the prompt that it can remind us of that.
Diana (23m 42s):
Airtable AI is pretty user friendly. You know, once you get Airtable down, adding the AI is pretty easy. I was surprised at how easy it was to to work with it. And I think it is just the beginning of seeing what we can do with artificial intelligence to help us with our work and anything that can help us with writing about DNA is fabulous.
Nicole (24m 5s):
Agreed. I think writing is so valuable and often we don’t do it because we don’t have time. And so if that’s a roadblock for us, then using AI to help us get started with our writing and get something put to paper can be really valuable. So I hope that you’ll try it out. And if you’re using Airtable, and especially if you already have the Airtable paid plan, you should definitely try using AI tools. So I hope you enjoyed this episode and we’ll talk to you guys again next week. Bye
Diana (24m 34s):
Bye-Bye
Nicole (24m 33s):
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.
Links
Using Airtable AI Tools to Evaluate DNA Matches – https://familylocket.com/using-airtable-ai-tools-to-evaluate-dna-matches/
RLP with DNA 4.0 (2024) Research Log – https://www.airtable.com/universe/expy4V9HzRUxtJLvh/rlp-with-dna-research-log-2022
Writage – https://www.writage.com/
Enable Markdown in Google Docs – https://support.google.com/docs/answer/12014036
Research Like a Pro with AI Workshop – https://familylocket.com/product/research-like-a-pro-with-ai/
Sponsor – Newspapers.com
For listeners of this podcast, Newspapers.com is offering new subscribers 20% off a Publisher Extra subscription so you can start exploring today. Just use the code “FamilyLocket” at checkout.
Research Like a Pro Resources
Airtable Universe – Nicole’s Airtable Templates – https://www.airtable.com/universe/creator/usrsBSDhwHyLNnP4O/nicole-dyer
Airtable Research Logs Quick Reference – by Nicole Dyer – https://familylocket.com/product-tag/airtable/
Research Like a Pro: A Genealogist’s Guide book by Diana Elder with Nicole Dyer on Amazon.com – https://amzn.to/2x0ku3d
14-Day Research Like a Pro Challenge Workbook – digital – https://familylocket.com/product/14-day-research-like-a-pro-challenge-workbook-digital-only/ and spiral bound – https://familylocket.com/product/14-day-research-like-a-pro-challenge-workbook-spiral-bound/
Research Like a Pro Webinar Series 2024 – monthly case study webinars including documentary evidence and many with DNA evidence – https://familylocket.com/product/research-like-a-pro-webinar-series-2024/
Research Like a Pro eCourse – independent study course – https://familylocket.com/product/research-like-a-pro-e-course/
RLP Study Group – upcoming group and email notification list – https://familylocket.com/services/research-like-a-pro-study-group/
Research Like a Pro with DNA Resources
Research Like a Pro with DNA: A Genealogist’s Guide to Finding and Confirming Ancestors with DNA Evidence book by Diana Elder, Nicole Dyer, and Robin Wirthlin – https://amzn.to/3gn0hKx
Research Like a Pro with DNA eCourse – independent study course – https://familylocket.com/product/research-like-a-pro-with-dna-ecourse/
RLP with DNA Study Group – upcoming group and email notification list – https://familylocket.com/services/research-like-a-pro-with-dna-study-group/
Thank you
Thanks for listening! We hope that you will share your thoughts about our podcast and help us out by doing the following:
Write a review on iTunes or Apple Podcasts. If you leave a review, we will read it on the podcast and answer any questions that you bring up in your review. Thank you!
Leave a comment in the comment or question in the comment section below.
Share the episode on Twitter, Facebook, or Pinterest.
Subscribe on iTunes, Stitcher, Google Podcasts, or your favorite podcast app.
Sign up for our newsletter to receive notifications of new episodes – https://familylocket.com/sign-up/
Check out this list of genealogy podcasts from Feedspot: Top 20 Genealogy Podcasts – https://blog.feedspot.com/genealogy_podcasts/
Leave a Reply
Thanks for the note!