Open GeneaGPT (the community-built genealogy AI tool) has been updated from version 2 to version 3 as of Monday 22 January 2024. This is a significant update, transforming Open GeneaGPT into a smarter and more engaging companion for exploring family history, making it easier and more enjoyable for everyone. It brings changes like better conversations, more helpful suggestions for next steps, ensuring a friendly and insightful journey into the past.
OpenAI’s custom GPTs are amazing tools for creating specialized bots that can perform various tasks, such as extracting data; describing documents, images, records; generating reports, stories, images; and much, much more. These GPTs are tools for the way you work; you build the AI tool you need, or find one on the store shelf. I have been exploring genealogy prompt engineering over the past year, building genealogy GPTs since November, and I have learned a lot about how to create and use them effectively. Recently, OpenAI launched their GPT Store, where users can share their custom GPTs with other ChatGPT Plus subscribers. This is a great opportunity to discover new bots and learn from others. I have shared five or six of my custom genealogy GPTs in the store, which are related to family history research (more are being finished in the lab). Custom GPTs are available to ChatGPT Plus subscribers ($20/month) for no additional cost in the ChatGPT Store.
These are the kinds of genealogy AI tools that I teach students to create; no programming skills are necessary. ChatGPT can interview you, to ask you what kind of AI tool you’d like to create. Then I can show you how to fine-tune the tool to suit your exact genealogy workflow. It won’t fetch you a Snickers Bar or promise to solve a 120-year family history mystery, but within the realm of what large language models can do today, this one does okay.
If you’d like hands-on, step-by-step instruction on how to build your own custom GPTs and other specialized genealogy AI tools, your own flock of bots, the National Genealogical Society is now enrolling for Prompt Engineering and Specialized AI Tools for Genealogists, which starts at the end of January 2024; six hours of instruction over four weeks is just the start; also includes a collaborative study group, sharing successes and learning from failures. The first section of 50 seats sold-out in days, so a second section has been opened. Learn more here: https://www.ngsgenealogy.org/ai/
ChatGPT with Vision (GPT-4V) analyzes a handwritten WWII draft card, and not only reads handwriting correctly but also accurately identifies text fields (name, address, next of kin, date of birth, occupation, etc.). This draft card, from my maternal grandfather Dean Lawrence (1921-2003) was chosen for the average block print handwriting; testing will continue to determine how good GPT-4V is with recognizing cursive handwriting, but it appears to handle average print well. Again, though, the innovation is that the AI will correctly label the data; that is, earlier OCR can process simple block handwriting, but it couldn’t tell you whether the text was name, address, or occupation. ChatGPT with Vision (GPT-4V) correctly labels the data. Now, imagine a simple Python script that processes a folder of 20 or 200 or 2000 images like this for you; that will be a trivial task once OpenAI makes API access available for GPT-4V, which is expected (more simply, that is, with middle-school programming skills, tasks such as these will be quickly and easily automated).
PROMPT: You are a professional genealogist; your area of expertise is documents and records; you are adept at extracting every piece of information from a document or record. You are also a skilled writer and image analyst. Think now about a World War II Draft Registration Card, also called DDS Form 1: list every conceivable piece of information that might be found on a draft card (focus especially on names, relationships, dates, places, events, and facts).
Then, find attached a draft card. Extract every piece of information from the image of the draft card.
Finally, create a biographical narrative report from the text on the draft card; dry recitation, no editorializing.
A couple of points to highlight:
you can assign multiple roles simultaneously to the AI: you can prompt the AI to act not only as an expert genealogist, but at the same time to act as a skilled writer and image analysist.
you can ask for several tasks, essentially stepping ChatGPT through a process.
you can ask for a narrative summary of the text in the image, e.g., “create a biographical narrative report from the text on the draft card; dry recitation, no editorializing.”
Here was the response:
RESPONSE:
Information Extracted from the Attached Draft Card: 1. Serial Number: 234 2. Full name of the registrant: Warren Dean Lawrence 3. Place of residence: West Jefferson, Ashe, NC 4. Mailing address: Same as residence 5. Telephone number: Not provided on this card 6. Age in years: 20 7. Date of Birth: May 21, 1921 8. Place of Birth: West Jefferson, USA 9. Name and address of someone who will always know the registrant's address: Mrs. Warren Dean Lawrence, West Jefferson, NC 10. Employer's name and address: Partially illegible but appears to reference a mining company 11. Place of employment or business: West Jefferson, Ashe, NC 12. Signature of the registrant: Appears to be "Dean Lawrence" or a similar variant 13. Date of registration: Not clearly visible on this card
Biographical Narrative Report: Warren Dean Lawrence, residing in West Jefferson, located in Ashe County, North Carolina, registered for the World War II draft. Born on May 21, 1921, in West Jefferson, USA, he was 20 years old at the time of registration. His mailing address was the same as his place of residence. The individual who would always be aware of Warren's address was Mrs. Warren Dean Lawrence, also of West Jefferson. Warren's employment details reference what appears to be a mining company located in West Jefferson, Ashe, North Carolina. The draft card carries what seems to be his signature, "Dean Lawrence" or a similar variant.
Prompt Chain to process data further
Once we have the “Information Extracted from the Attached Draft Card,” that information can be standardized, so that if you have many sources of this type (in this example, a draft card), the data can be collected, compiled, compared, and analyzed. For example, if you had 20 of these cards, you may wish to move that extracted data into a spreadsheet or database. Saving the data for each record as a CSV (Comma Separated File, a widely-used file format) file allows us to do this simple task. In the instance, I followed-up the first prompt with a second prompt; using the results of one task as the input for a subsequent task is called prompt chaining. Here was my second prompt:
PROMPT: Create a CSV file from the Information Extracted from the Attached Draft Card. Use the format: "FIELD NAME","VALUE", where FIELD NAME is the field label used in the original image and VALUE is the value extracted from the image, such as "NAME","Warren Dean Lawrence". Display the response in a code window for easy copy-and-paste.
ChatGPT responds with the data in the format of a CSV file, and places that information in a code window, which makes copying-and-pasting easy (just click the “📋 Copy code” icon). One beta mode of ChatGPT, called “Advanced Data Analysis” (formerly “Code Interpreter”), will create an actual CSV file and present you with a download link, but for now we can only use one beta feature at a time; expect that to get better in time.
Up Next: Clean, simple cursive handwriting on a draft card
There remain many document types to test. I’ve been testing more challenging samples to discover where the limit of ChatGPT’s handwriting recognition. Up next is a clean, simple cursive script on a form such as this draft card. Ultimately, however, we will test handwritten documents such as letters, diaries, journals, court orders, and probate files. Handwriting recognition is hard, so I expect the limit might be discovered sooner than anticipated, for the time being.
For the record, here is a screenshot of the beginning of this ChatGPT conversation:
The post was prepared as part of my upcoming talk: The National Genealogical Society is hosting a MemberConnects! event on Empowering Genealogists with Artificial Intelligence. Join us Wed. 6 Sept 2023, at 8 p.m. ET. Register here: https://bit.ly/NGSMemberConnects6Sept2023
As we venture deeper into the digital age, the intersection of genealogy and artificial intelligence (AI) is becoming increasingly more exciting. Today, I’m diving into a key skill that will empower your genealogical research with AI: how to craft effective prompts for large language models (LLMs) like ChatGPT.
If you’ve been following the ongoing conversation about genealogy and AI, you’ll recall that while LLMs have an arsenal of abilities, they also come with some quirks. To guide them effectively, we need to be thoughtful in our approach.
So, without further ado, let’s discuss the five components that make for a great genealogy prompt!
1. A Role: Begin by imagining you’re recruiting an expert for a specific task. What’s their profession? What expertise should they possess? By defining a role, you’re setting the stage and giving the LLM a context to operate within.
2. A Goal: What do you want to achieve? This is your endgame. Being explicit here ensures that the AI has a clear understanding of your expectations.
3. A Text: Since genealogy is grounded in factual research, provide the text you’d like the AI to process. By supplying the exact text, you mitigate the risk of the AI generating fictitious information (a phenomenon known as “hallucinating”).
4. A Task: Simplify your request. Break it down into manageable steps, just as you’d explain a process to a bright but inexperienced intern. This ensures the LLM knows the sequence of operations it should perform.
5. A Flask: While the term “flask” is playfully chosen for the rhyme, it represents the container or format you’d like your response in. This could range from a narrative report to a structured database table.
For instance, consider this prompt:
PROMPT: You are an expert genealogist and a seasoned data scientist. Your goal is to extract structured relationship data from a provided obituary. Find below an obituary for John Smith from the July 20, 1923, New York Times, page 18. Extract any explicitly stated relationship information; include a quote that supports that determination. Format your response as a CSV file and display that in a code window for easy copy-and-paste.
By following this structure, you provide the LLM with a roadmap to navigate your request. The result? You get precisely the information you’re looking for, formatted just the way you want it.
In conclusion, the digital world of genealogy is at our fingertips, and with tools like ChatGPT, the possibilities are endless. But remember, like any tool, its effectiveness lies in how we use it. With a role, goal, text, task, and flask in hand, you’re all set to harness the full power of LLMs for your genealogical pursuits. Happy researching! 🌳🔍🖥️
I. Introduction: The Power of A.I. Genealogical Prompt Chaining
Welcome to the fascinating world of AI applications, where the dynamic use of language models continually expands our capabilities and insights. Today, we venture further into this realm to explore an exciting concept: prompt chaining.
Prompt chaining is a powerful tool that taps into the versatility of AI, enabling us to extend the utility of a single interaction by linking it to subsequent prompts. In essence, it is a relay race of information processing where the output of one AI operation serves as the input to another. This chain of operations harnesses the potential of AI to process, generate, and analyze data in a seamless flow, creating opportunities for multi-faceted inquiries and sophisticated data manipulation. Most simply, prompt chaining is using the response from one A.I. interaction as the prompt for a following interaction.
Our exploration of prompt chaining builds upon the foundations we established in previous blog posts. We delved into structured data extraction from narrative sources, shedding light on how we can extract valuable insights from complex texts. In another post, we introduced ChatGPT plugins, an innovative extension of OpenAI’s powerful language model, GPT-4, which further enhances its capabilities.
Today, our journey takes us to a fascinating intersection of these concepts. We aim to illustrate how prompt chaining can be effectively combined with the “Show Me” diagramming plugin for ChatGPT. This plugin translates textual relationships into illustrative diagrams, providing a visual representation of the relationships extracted from narrative sources.
However, a quick but important caveat before we proceed: the advanced features we’ll be discussing, including the “Show Me” plugin, require a subscription to ChatGPT Plus. While the primary value of this subscription lies in gaining access to the latest and most powerful iteration of the language model—GPT-4—the plugins definitely add a delightful layer of functionality, the proverbial icing on the cake.
This post will introduce and develop an understanding of prompt chaining. Building on a general definition and description of prompt chaining, four cases studies will be considered to sharpen our imagination of how prompt chaining will evolve as the number of A.I.-assisted genealogical use cases (tasks) grows.
Most simply, prompt chaining is using the response from one A.I. interaction as the prompt for a following interaction.
A.I. Genealogy Insights
II. Background Information
The “Show Me” diagramming plugin is an innovative tool integrated with OpenAI’s language model, ChatGPT. It’s designed to help generate visual diagrams based on user inputs. With this plugin, users can request ChatGPT to draw diagrams, flowcharts, or even family trees. Its applications are broad, making it a versatile and practical tool for visual learners, project managers, educators, and in our case, genealogists.
To fully appreciate the potential of the “Show Me” diagramming plugin and how it can be utilized in genealogical research, it’s essential to familiarize oneself with some of our previous posts. These articles delve into the practical applications of AI in genealogy, and specifically, how ChatGPT can be utilized to extract and organize valuable information.
The article “AI and Genealogy: Using ChatGPT to Glean Info from Obits, Articles, and Announcements” provides a comprehensive exploration of how ChatGPT can extract genealogical information from text sources like obituaries, wedding announcements, and newspaper articles. The post demonstrates how the AI can be prompted to identify relationships between individuals, create markdown tables of this information, and even generate GEDCOM files, which can be used to create family trees.
Another key post to review is “Using ChatGPT Plugins for Genealogy“. Leveraging ChatGPT Plus plugins can substantially enhance artificial intelligence-assisted genealogy. These software add-ons customize ChatGPT for genealogical research, with the AskYourPDF plugin as a key example. This tool allows for interaction with PDF files, extracting crucial data from scanned records or historical articles. These plugins automate tasks and interact with digital resources, increasing efficiency and insight in genealogical research, symbolizing the promising future of AI in genealogy. The plugin used extensively in this post, the “Show Me” diagramming plugin, is also introduced in that earlier blog post.
Those two previous posts provide a solid foundation for understanding how AI, and more specifically, ChatGPT with its plugins, can revolutionize genealogical research. As we move forward today, we will explore how the “Show Me” diagramming plugin, in particular, can provide a visual dimension to relationships.
You know the saying, “A picture is worth a thousand words.” That saying is never more true than in genealogy–relationships are hard for others to understand when presented only as a stream of names. Seeing a diagram of the relationships among people–a family tree or a pedigree chart–is a tremendous aid to understanding. Visually presenting relationships helps us communicate our research and findings in a way that others are more easily able to understand. The “Show Me” plugin empowers genealogists to use A.I. to generate charts, diagrams, and family trees from textual, natural language descriptions.
The advent of A.I. plugins, however, in early summer 2023 is just developing from infancy to early childhood. My research and experience during this season is that many plugins have great potential, but currently exhibit significant shortcomings. Specifically, the “Show Me” plugin demonstrated extensively in this post would currently received a grade of “C-” but with the expectation that this will, for reasons I will explain below, improve soon to a useful “B+”.
Finally, before we jump into the case studies, a brief preview and explanation. The last two case studies involve research as genealogists might typically conduct: the consideration of an earlier researcher’s work and the examination of an obituary. The first two case studies, however, use as their subject matter fictional places and people. This was done for several reasons. First, readers are likely to be familiar with one or both of the peoples and places in these first two case studies. Second, the people and relationships offered by these fictional examples are extensive, allowing for a better presentation of the principles being demonstrated. And finally, third, well, it was fun. I pray these first two fictional examples won’t distract any readers so much as to inhibit their ability to see the underlaying processes at work. That said, let’s jump into the case studies.
III. Case Study 1: Complex Family in “One Hundred Years of Solitude”
“One Hundred Years of Solitude,” a widely acclaimed novel by Gabriel García Márquez, offers a fascinating exploration of the Buendía family over seven generations in the fictional town of Macondo. The intricate relationships and complex family dynamics in this novel make it an ideal case study for demonstrating the capabilities of ChatGPT and the “Show Me” diagramming plugin in handling complex genealogical data.
To begin, we would need to extract the relationship information from the text. This process involves identifying and understanding the connections between the various characters in the novel. With the assistance of ChatGPT, this is made significantly easier. By providing the AI with a well-crafted prompt, we can instruct it to sift through the text and identify the relationships between characters, as well as the events that bind them together.
For example, a prompt may be constructed as follows:
PROMPT: Assume the role of an expert, professional genealogist. Consider the genealogically relevant information that might appear in the novel "One Hundred Years of Solitude". I would like to know about the stated relationships between characters in the text. When you can with certainty, state the relationship between two characters in the book. Extract names and relationships from "One Hundred Years of Solitude" and present that information in a way that the next AI, a diagram-generating AI, can use your information, in turn, to create a family relationship visualization.
ChatGPT, with its impressive language processing capabilities, then extracts the relationship information and organize it. [NOTE: In this example, I did not feed the entire novel to ChatGPT to process. In all the following examples and case studies, however, the A.I. is restricted to using only information that I provide to it.]
The response did not look enough like a prompt for the next AI, so I prompted ChatGPT to refine its response. As we have noted before, just as “writing means re-writing,” so working with A.I. is often an iterative, refining process; that is, we don’t often get things perfect on the first attempt, so try, try again, until you get what you need.
The refinement looked like a great prompt for the “Show Me” plugin. So, once we have this relational data, we can use it to create a prompt for the “Show Me” plugin to generate a visual representation of the Buendía family tree. The next prompt, then, could look something like this, which is just an exact copy-and-paste from the last response:
"SHOW ME" PROMPT: Show me a family tree for the Buendía family from "One Hundred Years of Solitude" by Gabriel Garcia Marquez using the following relationships:
1. José Arcadio Buendía is married to Úrsula Iguarán. They have two children: Aureliano Buendía and José Arcadio.
2. Aureliano Buendía is married to Remedios Moscote. They have a daughter named Renata Remedios, also known as Meme.
3. José Arcadio is married to Rebeca. They have a son named Arcadio.
4. Arcadio is married to Santa Sofía de la Piedad. They have three children: Remedios the Beauty, Aureliano Segundo, and José Arcadio Segundo.
5. Aureliano Segundo is married to Fernanda del Carpio. They have three children: Renata Remedios (also known as Meme), José Arcadio (II), and Aureliano (II).
6. Meme is married to Mauricio Babilonia. They have a daughter named Renata Remedios, also known as Renata.
7. José Arcadio (II) is married to Renata. They have a son named Aureliano (III).
Please note that there are multiple characters named Aureliano and José Arcadio in the family, so make sure to differentiate between them in the family tree.
The result would be a detailed diagram illustrating the family relationships of the Buendía family. This visual representation, built on the foundation of meticulous data extraction and AI processing, allows us to easily understand and analyze the complex familial dynamics present in “One Hundred Years of Solitude”.
Sharp-eyed readers will here observe why I currently graded the “Show Me” plugin as a “C-“. Notice that spouses are not placed on the same row or level as their partners, but instead are one level below with the children. There are work-arounds and solutions to this problem (some easy, as noted below), but users should be aware of this issue in the spring/summer of 2023. Both the “Show Me” plugin and the other diagramming plugins available in the ChatGPT plugin store rely on a graphic tool called “Mermaid“; Mermaid is great for many, many uses, but it is, obviously, not created primarily to generate family trees. The developers of Mermaid, though, are working to explicitly add family tree diagrams to the tool’s repertoire. As I mentioned, though, there are work-arounds currently available, and every diagram that is generated is easily editable with a link that is included with every chart.
So, that shortcoming acknowledged, do not let the perfect be the enemy of the good. What ChatGPT and the “Show Me” plugin did, even though not perfect, is still fairly amazing. In about three-and-a-half minutes, the genealogically relevant information was extracted from a 440 page novel and transformed into a family tree. The implications of this process are significant. Not only does it provide a visual aid for understanding complex relationships in literature, but it also has the potential to be a valuable tool in fields such as education, history, and of course, genealogy. By making it easier to visualize and comprehend complex familial relationships, the “Show Me” plugin enhances our ability to extract deeper meanings and insights from texts. Moreover, it demonstrates the potential for AI to transform how we approach, engage, and understand large bodies of text.
IV. Case Study 2: Tolkien’s Universe
The second case study, as I mentioned, also draws from the world of relationships in literature. I’m no Tolkien-head, so it was a bit unusual for a “Lord of the Rings” post to appear in my social feed. I’ve enjoyed the movies and read a couple of the books, and I know just enough to have heard that Tolkien is often credited with developing a literary universe. So when glanced at the article in my feed, I quickly noticed that it would lend itself nicely to this demonstration of chain prompting data extraction and data visualization. Tolkien’s universe offers a compelling backdrop against which we can further demonstrate the potential of prompt chaining with ChatGPT and the “Show Me” diagramming plugin.
The article I stumbled across was rich in relationships between people and, well, something else. So, whether you’re navigating the familial ties of the House of Elrond or tracing the lineage of the Dwarven clans, Tolkien’s works are filled with complex genealogical data that could be readily extracted by an AI. To begin this process, we would again need to construct a suitable prompt that instructs ChatGPT to identify and delineate the relationships between the various characters within Tolkien’s universe.
A sample prompt might look like this:
PROMPT: You are a professor of literature with deep expertise in science fiction and fantasy. Find below an article about Tolkien's "Lord of the Rings" universe. As if you were a genealogist, I want to map, chart, diagram, or in some way visualize the relationships between creatures and characters in the article. Please suggest how that might BEST be done, then attempt to do whatever you can do to visualize the relationships, even if it's not the best possible way imaginable.
ARTICLE:
Lord of the Rings: What Does Sauron Look Like Under His Armor? It's Complicated
https://www.cbr.com/lord-of-rings-what-sauron-really-looks-like/
Guided by this prompt, ChatGPT would then delve into the texts and extract the necessary relationship data, organizing it into a structured format for easy interpretation and analysis.
With the relationship data in hand, we can then feed this information into the “Show Me” plugin to generate a diagram that visualizes these connections. This might involve creating a new prompt such as:
"SHOW ME" PROMPT: Show me a diagram of the lineages and relationships in Tolkien's universe using the following relationship data.
TEXT:
I gathered a lot of information on the cosmology and beings of Tolkien's universe, which can serve as a basis for creating a diagram or chart to visualize the relationships between the creatures and characters in the "Lord of the Rings". Here are the main points:
[COMPLETE DESCRIPTION FOLLOWED]...
The result? A comprehensive and easily digestible diagram that visually represents the complex relationships in Tolkien’s universe. This visual aid can be a powerful tool for understanding and exploring the intricacies of Tolkien’s rich lore, making it accessible and engaging for both seasoned fans and newcomers alike. It should not take too much imagination to see how this process is analogous and applicable to genealogy.
Furthermore, this case study underscores the potential of prompt chaining with AI in enhancing our understanding of literature. By enabling us to visualize complex relationships and lineages, the “Show Me” plugin can bring to life the rich tapestry of connections that form the backbone of our favorite family history stories. In this way, we can unlock deeper insights and appreciation for the families that we love, showcasing the immense potential of AI as a tool for genealogical analysis and exploration.
V. Case Study 3: Family Biographical Sketch
As we transition from literature to a genealogical context, let’s move into the realm of family biographical sketches. These sketches are rich sources of family history, often containing detailed information about relationships, life events, and personal narratives that can be invaluable in genealogical research.
In this case study, we’ll explore how a family biographical sketch can provide the data needed for prompt chaining and how this approach can bring the intricate web of familial relationships to life. In this example, it took me about six minutes to take a photo of a book page with my phone, extract the text from the phone, then use ChatGPT 4 to extract the relationship information from the text, and finally have the “Show Me” plugin generate a family tree diagram.
Family biographical sketches typically provide in-depth information about a family, including relationships, dates and locations of important events, occupations, and sometimes even anecdotes that give unique insights into the lives of our ancestors. These sketches, while rich in information, can sometimes be challenging to navigate due to their narrative format and the sheer volume of information they contain.
Here’s where prompt chaining with ChatGPT and the “Show Me” plugin can prove invaluable. By directing ChatGPT to extract the genealogical data from the biographical sketch, we can transform the narrative data into a structured format, ready for visualization.
The initial prompt could be framed like this:
PROMPT: You are an expert genealogist and the world's best prompt engineer. Find below an excerpt from a genealogical text. Your response will be passed to another AI which is capable of creating a diagram of family relationships. Extract information about people and relationships. Interpret that information and use it to craft a prompt for the diagramming AI. Your response should be in the form of a prompt for the other AI.
EXCERPT:
THE AMBROSE PARKS LITTLE FAMILY
The earliest known ancestor of the numerous Littles in Ashe County is Isaac (ca. 1800-1892).
[BIOGRAPHICAL SKETCH CONTINUES...]
The results…
Once the relationships are extracted, the data can be fed into the “Show Me” plugin with a prompt like this:
"SHOW ME" PROMPT: Based on the provided genealogical text below, show me a diagram representing the reported relationships. Combine Husbands/Wives into one node; for example "Isaac Little married Elizabeth Poe" is one node, and the parents together of their children.
GENEALOGY:
1. The earliest known ancestor is Isaac Little (ca. 1800-1892), who married Elizabeth Poe, daughter of Mathias and Sarah (Grimsley) Poe. Isaac and Elizabeth had thirteen children, including Ambrose Parks Little.
[GENEALOGY CONTINUES AS SHOWN ABOVE...]
The resulting diagram brings clarity to the complex relationships described in the biographical sketch. This visual representation of family connections can be an invaluable tool for genealogists, making it easier to understand and trace familial relationships across generations. Note the instruction in bold in the prompt; this is the work-around to get the “Show Me” plugin to place mothers and fathers on the same row, line, or level in a family tree diagram; for now, they need to be treated as one unit, that is, as the parents together. Again, this deficiency will be addressed and remedied in time. Advanced users may want to note that the Mermaid scripting language is easily accessible and configurable for every diagram that the “Show Me” diagram generates.
Furthermore, it highlights the power of prompt chaining in a genealogical context. By transforming narrative data into structured, visual representations, we can unlock a new dimension of understanding in our family history research. This case study serves to showcase how the combined power of AI and genealogy can enrich our understanding of our past, bringing our ancestors’ stories to life in a compelling and accessible way.
VI. Case Study 4: Obituaries
Continuing in our genealogical context, let’s turn our attention to obituaries, a commonly used but often underappreciated resource in genealogical research. Obituaries are a treasure trove of information, offering insights into the lives of our ancestors, their relationships, and their place in the community.
In this case study, we’ll explore how obituaries can provide valuable data for prompt chaining and how this method can bring new insights to genealogical research.
Obituaries are unique in their ability to encapsulate the essence of a person’s life, including their family ties, accomplishments, and legacies. However, the narrative format and sometimes cryptic language used in obituaries can make extracting this information a challenge. This is where AI, particularly ChatGPT, can step in.
To initiate the process, we would provide ChatGPT with a prompt to extract relevant information from an obituary. The prompt could be framed as:
PROMPT: You are an expert genealogist and the world's best prompt engineer. Find below an excerpt from a genealogical text. Your response will be passed to another AI which is capable of creating a diagram of family relationships. Extract information about people and relationships. Interpret that information use it to craft a prompt for the diagramming AI. Your response should be in the form of a prompt for the other AI.
OBITUARY:
Mrs. Houck, 53, Buried Saturday
Funeral service was held on Saturday at 11:00 a.m., at the Bethany Church, for Mrs. Annie Houck, 53, of West Jefferson, who died at her home on Thursday, January 20, after being ill for some time. The service was conducted by Rev. R. G. Wagoner and Rev. T. J. Houck, and burial followed at the Bethany Cemetery.
Mrs. Houck, who was a native of Ashe County, the daughter of Jacob Goodman and Sarah Tatum Goodman, is survived by her husband, James Houck, and the following children: Mrs. Dora Howell, Todd; Fay, Agnes and Jessie D. Houck; Phil Houck, of the U.S. Army, stationed in South America, and J. C. Houck, Jr., of the Navy, stationed in Maryland.
The data extracted by ChatGPT can then be used to create a prompt for the “Show Me” plugin to visualize these relationships:
"SHOW ME" PROMPT: Based on the provided genealogical text below, show me a diagram representing the reported relationships. Combine Husbands/Wives into one node; for example "Annie Goodman married James Houck" is one node, and the parents together of their children.
Based on the provided genealogical text, please create a diagram representing the following relationships:
1. Mrs. Annie Houck, born as Annie Goodman, was the daughter of Jacob Goodman and Sarah Tatum Goodman. She was married to James Houck.
2. Annie and James had six children: Mrs. Dora Howell; Fay Houck; Agnes Houck; Jessie D. Houck; Phil Houck, who is in the U.S. Army, stationed in South America; and J. C. Houck, Jr., who is in the Navy, stationed in Maryland.
Please represent these family relationships, including marriages and offspring, in the family diagram.
The resulting diagram provides a visual representation of the relationships gleaned from the obituary, offering a unique perspective on the family structure and relationships of the deceased.
The value of this approach in genealogical research is manifold. Not only does it allow for the efficient extraction of valuable information from obituaries, but it also presents this information in a clear, visual format that can aid in understanding complex family relationships and in tracing lineages. It demonstrates the power of prompt chaining in unlocking the potential of obituaries as a source of genealogical data, bringing us one step closer to the stories of our ancestors.
VII. Conclusion
Looking forward, the potential of prompt chaining is immense. As the capabilities of AI tools like ChatGPT continue to expand, the possibilities for prompt chaining grow exponentially. The fine-tuning of the diagrams provided by the “Show Me” plugin, for instance, could lead to more nuanced and detailed visualizations of data, further enhancing our understanding of complex information.
Every week, we witness the introduction of new use cases – data extraction, translation, optical character recognition (OCR), narration, and diagramming, among others. With each new task that AI can handle, the potential combinations of prompt chaining multiply dramatically, broadening the horizons of AI-assisted endeavors, such as genealogy. This continuous growth in AI capabilities underscores the increasing relevance and power of prompt chaining.
In the present, we manually chain or link prompts, carefully curating the sequence to accomplish our specific goals. Yet, we can envision a future, perhaps not too distant, where AI agents will not only suggest optimal task combinations but also carry out complex workflows seamlessly, further reducing the manual effort and increasing the efficiency of achieving more intricate goals.
I invite all readers to dive into this fascinating world of AI-assisted research. Try out the “Show Me” plugin and experiment with prompt chaining. The journey may challenge your understanding of what AI can do and inspire you to imagine new ways to harness its power.
We would love to hear about your experiences, discoveries, and insights in this area. Please share your results in the Comments section below. Remember, every experiment, every question, every insight adds to our collective understanding of this powerful tool.
Here’s to the continued exploration of AI’s potential, and to the remarkable power of prompt chaining!
Humans vs. Machines with Gary Marcus 🤖: “And the winner is…Watson!” 🏆
Last Week in AI 🧠 (segment at 44:00): “Auto-GPT and BabyAGI: How ‘autonomous agents’ are bringing generative AI to the masses” 🤖
Video: Interview: Adam Conover: A.I. and Stochastic Parrots | FACTUALLY with Emily Bender and Timnit Gebru
Prompt: A Useful, Flexible Format
My most frequent initial prompt is fairly similar to this helpful graphic I saw on Twitter this week. My most-used prompts for various topics usually take the general form:
PROMPT: Assume the role of a professor of [field] with a specialty in [sub-field]. Find below a [text of some sort]. Create a [task] in the form of a [format].
My most commonly used response formats are markdown tables (which copy-and-paste nicely into word processors) and bullet lists (which are often a next set of prompts); lesser used are hierarchical outlines and code windows.
ChatGPT Cheat Sheet:
This is one of the easiest and most effective ways to use ChatGPT.
Great use case from Denys Allen, PA Ancestors. Love the emphasis on having ChatGPT *process* information rather than gathering information (which WILL burn you), and generating text based on *given* (NOT gathered) information.
Understanding the difference between instructing chatbots to PROCESS information rather than GATHER information is vital in spring 2023 to their successful use. Too frequently we inadvertently ask ChatGPT to gather, find, or search for information without realizing that doing so is inviting the large language model to inject fiction and hallucinations into the response, because that is their nature. The antidote is to constrain and restrict the chatbot by instructing it to work ONLY on the information you provide it.
I've used ChatGPT to write family history narratives in less than five minutes. Here's the steps I used.
— Denys Allen, PA Ancestors (@denys_allen) April 24, 2023
Use Case: Historical Document Analysis and Simplification
In the Genealogy and Artificial Intelligence group at Facebook, Rand Hall shared an interesting use case: Historical Document Analysis and Simplification. In this use case, the AI system is tasked with analyzing and summarizing a historical document, converting it into contemporary standard English, and comparing different summaries by extracting salient points. This involves understanding historical language and context, as well as transforming the text while maintaining its original meaning. The AI system’s capabilities can assist genealogists and researchers in interpreting, comparing, and understanding historical texts more effectively and efficiently.
PROMPT: Assume the role of a historian, linguist, and editor. Find below an 1882 Letter to the Editor. First, summarize the Letter. Then, rewrite the Letter in simple contemporary standard English, while prioritizing fidelity to the meaning of the original Letter.
Podcasts: Two Recommendations from This Week
While mowing the lawn this week, I listened to several great podcasts; these are the best two. The first, an episode of a new show from a familiar host, and the second, the best segment this week, a short introduction to “autonomous agents.”
Steve’s Note: Polished and produced, this podcast is beginner and novice-friendly, and serves an antidote to the hype and hustle surrounding AI today; this is the first episode of a limited series. Summary: The first episode of the fourth season of Humans vs Machines podcast is titled “And the winner is…Watson!” hosted by Gary Marcus. The episode talks about IBM’s Watson’s defeat of Ken Jennings on Jeopardy! and how it became one of artificial intelligence’s most dramatic triumphs. The podcast also discusses AI’s impact on humanity and its potential to change the world. The episode features interviews with David Ferrucci, an AI researcher who led the Watson project, and Ken Jennings of Jeopardy!
2️⃣ Podcast: Last Week in AI 🤖🧠 Episode: #119: Open Source GPTs, X.AI, Auto-GPT, China’s Censorship of AI Segment (timestamp: 44:00 to 53:00): Auto-GPT and BabyAGI: How ‘autonomous agents’ are bringing generative AI to the masses URL:👉 https://lastweekin.ai/p/lwiai-podcast-119-open-source-gpts
Steve’s Note: This podcast is closer to the classic two techies talking format; I found the hosts very smart and informative; this 9-minute segment is the most interesting I heard this week, introducing a topic that is likely to become more important in the weeks and months to come; you can try one of these autonomous agents at https://agentgpt.reworkd.ai/. Summary: Autonomous agents are software programs that utilize large language models like GPT-4 to automate and simplify tasks such as research, code writing, and business management. Notable examples include BabyAGI and Auto-GPT, which offer various features and functionalities. Despite their potential, these agents face challenges in maintaining focus, predictability, safety, and reliability. They also raise ethical and social concerns regarding AI operating without human supervision. Nevertheless, autonomous agents represent progress towards artificial general intelligence (AGI), where AI systems can think and act like humans.
Video: Interview: Adam Conover: A.I. and Stochastic Parrots | FACTUALLY with Emily Bender and Timnit Gebru
Large language models are unmoored from reality, so I appreciate experts who can offer a grounding perspective.
Adam Conover, host of “Adam Ruins Everything,” may not be everyone’s cup of tea, but I personally enjoy his work. In this excellent hour-long podcast/YouTube video, he interviews two important AI experts: Emily Bender and Timnit Gebru. The conversation doesn’t present a balanced, centrist perspective, but it offers a valuable skeptical viewpoint. This is a great listen while doing other things, like driving or folding laundry, but it also deserves your full attention if you are deeply interested in the topic.
Any time a linguist such as Emily Bender gets a spotlight, I cheer, and ten-fold for computational linguists. Timnit Gebru was famously fired from Google for too loudly calling attention to bias in their training data and systems. I hope these voices are more widely heard.
