Category: News

“Man as Toolmaker” Is Taking on New Meaning at the Dawn of the AI Age

Kevin Borland introduced Linka, an AI assistant he trained. She writes code to improve features on his genealogy site, Borland Genetics, making it easier for users to share their family history with relatives who aren't members of the site.

About 14 months ago, I suggested genealogists would soon have a “flock of bots” assisting them. Borland Genetics is already making this a reality and even surpassing my expectations. Kevin Borland is creating at the cutting edge, orchestrating multiple AI assistants like a symphony conductor, pushing the boundaries of what’s possible. Yesterday, Kevin introduced Linka, an AI assistant he trained. Linka writes code to improve features on his genealogy site, Borland Genetics, making it easier for users to share their family history with relatives who aren’t members of the site.

But here’s what’s truly exciting: ALL genealogists can start creating their own helpful assistants, even at a basic level. Andrew Redfern’s upcoming webinar on saved prompts is a perfect example of how accessible this technology has become. Saved prompts are a basic AI skill that saves time by allowing users to re-use elaborate or simple prompts developed over time, with digital resources where you need them, for when you need them. The products have different names at different venders: Custom GPTs at OpenAI, Projects at Anthropic, and Gems at Google’s Gemini. While our creations may not initially match the sophistication of Kevin’s Linka, these assistants can still be amazingly useful and time-saving. And the pace of computational advancement is only accelerating—the next couple of years could be dizzying.

Saved prompts are a basic AI skill that saves time by allowing users to re-use elaborate or simple prompts developed over time, with digital resources where you need them, for when you need them.

AI Genealogy Insights

It’s exhilarating to witness this explosion of new tools created by everyday family historians alongside industry leaders. Technology like this is genuinely leveling the playing field, opening possibilities that once felt like science fiction. (I planned to share this post yesterday, when Kevin announced Linka, but held off to avoid it being mistaken for an April Fool’s Day prank—even though it sounds like science fiction, this is real.)

If you’re intrigued, don’t miss Andrew’s session at the “GPTs for Family History: Unlocking the Potential of AI” webinar during the 6th Annual 24-Hour Genealogy Webinar Marathon hosted by FamilyTreeWebinars.com and MyHeritage. The marathon begins Thursday, April 3 at 5pm Eastern U.S. time (Friday, April 4 at 8am Sydney time). Andrew’s webinar begins Thursday, April 3 at 8pm EDT.

Kevin Borland introduced Linka, an AI assistant he trained. She writes code to improve features on his genealogy site, Borland Genetics, making it easier for users to share their family history with relatives who aren’t members of the site.

This week marks the one-week anniversary of OpenAI’s improved image generator, GPT-4o with image generation. I’m guessing that the image of Linka that Kevin shared was generated by GPT-4o. This advance in image generation is marked by three characteristics: a leap in photorealism, almost perfect in-image text rendering (meaning it’s much better at incorporating words into images), and the ability to understand and implement long instructions in the image prompt. These new abilities are a manifestation of the new auto regressive image generation.

Sources:

Andrew’s webinar on saved prompts: https://familytreewebinars.com/webinar/gpts-for-family-history-unlocking-the-potential-of-ai/ https://familytreewebinars.com/24-marathon/

Borland Genetics: https://www.borlandgenetics.com/

Kevin Borland’s original introduction of Linka: https://www.facebook.com/kevin.borland.18/posts/pfbid0onRjUWSG91cM8NrhFFXjtsrVTj1ZzQYkSssWT6xiSjL5zZWSH2YaJT5X1fezg6wpl

“Linka” entry at International Society of Genetic Genealogy: https://isogg.org/wiki/Linka

How Are You Using AI?

📢 Friends, Family Historians, and Genealogists! How Are You Using AI? 🤖

Mark Thompson and I are excited to be collaborating with CeCe Moore and the Institute for Genetic Genealogy to present I4GG AI Day 2025 on Friday 28 March 2025, both in person in San Diego and virtually.

I4GG AI Day 2025 is a Full Day of Discovery, Instruction, and Learning:

Learn Transferable AI Skills with Family History examples

  • Five classes covering Foundations, Prompting, Research, Writing, Images
  • Expert panel on Responsible Use

Please take a moment to answer this quick poll so we can better tailor our sessions to your needs. 👇

How would you describe your current use of AI in genealogy?

  1. 🛑 No exposure – I’ve only read about AI or haven’t used it at all.
  2. 🟡 Still in the first 20 hours – I’ve dabbled and am just getting started.
  3. 🟢 Basic Literacy – I’ve used AI for far more than 20 hours and I feel comfortable and literate in their use, but I wouldn’t call myself an expert.
  4. 🚀 Power User – I feel proficient and understand best practices for AI in genealogy, such as differences between models and how to choose the right one, their limits, and how to mitigate those limits.

Drop your vote and feel free to comment on what AI tools you’ve tried! Looking forward to an engaging discussion at I4GG AI Day 2025! 🎤🔥

You can read more about the event and register to attend in person or virtually: https://i4gg.org

Quick Update on AI “Reasoning” Models as OpenAI Releases New o3 Variants

The AI-space has been ablaze with news about “reasoning” models since the release of DeepSeek’s reasoning model “R1” (which grabbed attention for its: 1) strength; 2) non-US origins; 3) open-source availability; and 4) cheap access). Now, as OpenAI releases their next reasoning models – “o3-mini” and “o3-mini-high” – it’s worth understanding what this “reasoning” buzz is all about. Here’s a higher-level approach and a more concrete example:

What’s Actually Happening Here?

Oversimplified, a “reasoning” model takes a few moments to refine its response before it gives you an answer. The Loathsome Jargon you may hear is “test-time thinking” or—the truly hideous Loathsome Jargon—”inference compute”. Functionally, it’s as if the model, before providing its response, asks itself if it understands the intent of your prompt, thinks things through “step-by-step,” and reconsiders its response and revises it, all before responding to your prompt. You experience this “reasoning” during use when you see the chatbot taking from a few seconds to a few minutes before responding to your prompt–seriously, reasoning models do not excel at back-and-forth conversation.

Current State of Play

OpenAI has launched two new AI reasoning models, o3-mini and o3-mini-high, pushing forward their capabilities in coding, science, and complex problem-solving while responding to competition from DeepSeek’s recent advances. The o3-mini model delivers responses 24% faster than its predecessor while maintaining comparable performance to o1, and notably, o3-mini is the first reasoning model available to free ChatGPT users. The o3-mini-high version offers enhanced performance for paid users, with both versions featuring adjustable reasoning effort levels (low, medium, or high) to balance speed and accuracy. Access to o3-mini allows unlimited usage for Pro subscribers ($200/month), while Plus and Team plan users ($20/$50/month) are limited to 150 messages per day (triple their previous limit), with Enterprise and Educational customers gaining access within a week. The model also introduces integrated search capabilities, providing up-to-date answers with web source links.

Free-tier users can now access OpenAI o3-mini, a specialized reasoning model balancing precision, speed, and efficiency. Optimized for technical domains, it offers enhanced accuracy with moderate reasoning effort—available via the ‘Reason’ option in ChatGPT. Experience advanced AI at no cost.

Because this sub-type of LLM is newer and less-used till now, their practical application is still being discovered, kinda like when GPT-4 first dropped in March 2023. In a nutshell, reasoning models complement rather than replace traditional LLMs, with different strengths and uses. Reasoning models are said to excel at PLANNING and ITERATIVE ANALYSIS. A perfect example appeared in “I Asked ChatGPT’s New ‘Reasoning’ Model to Craft a Research Plan–Here’s What Happened” (December 6, 2024), which tested o1’s planning capabilities through a complex genealogical research task. Notably, I departed from my usual practice of iterative chatbot conversation, instead testing the model’s initial, single-shot planning capability – exactly the kind of structured, thoughtful task where reasoning models shine1.

Practical Application

Really oversimplifying: Traditional models (ChatGPT GPT-4o, Claude 3.5 Sonnet, and Google Gemini) work best for most LLM processing tasks (summarization, extraction, generation, transformation). Reserve “o1”, “R1”, and now the new “o3” variants for when you’d actually like the model to spend some time on a more considered response, such as when asking for a plan, strategy, or analysis.

Using These Tools Effectively

TIPS: When using reasoning models be simple and direct with your goal or question WHILE providing as much context or detail as possible.

This is still early days with reasoning models, so expect new uses and best practices to be developed over time. For deeper insights, Prof Ethan Mollick’s recent work on reasoning models provides excellent context – particularly his December 2024 summary “What just happened“2 and his earlier introduction to reasoning models from September 2024, “Something New: On OpenAI’s ‘Strawberry’ and Reasoning.”3

A reasoning model isn’t a chat model—it’s a structured reasoning engine. Unlike GPT-4o, it won’t infer missing context, requiring users to input full details upfront. Treat it like a ‘report generator,’ not a chatbot, for more precise, higher-quality results.
Source: Nick Dobos, @NickADobos, https://x.com/NickADobos/status/1878267872079937637.

Footnotes

  1. Steve Little, “I Asked ChatGPT’s New Reasoning Model to Generate a Research Plan” – AI Genealogy Insights blog, https://aigenealogyinsights.com/2024/12/06/i-asked-chatgpts-new-reasoning-model-to-craft-a-research-plan-heres-what-happened/ (December 2024) ↩
  2. Ethan Mollick, “What just happened” – One Useful Thing blog (December 2024) ↩
  3. Ethan Mollick, “Something New: On OpenAI’s ‘Strawberry’ and Reasoning” – One Useful Thing blog (September 2024) ↩

I Asked ChatGPT’s New “Reasoning” Model to Craft a Research Plan–Here’s What Happened:

Readers interested in this post may also want to know about: "Quick Update on AI 'Reasoning' Models as OpenAI Releases New o3 Variants" (31 Jan 2025).

On Thursday 5 December 2024, the creators of ChatGPT kicked-off their “Twelve Days of OpenAI” by releasing the full version of their “reasoning” model, named “o1.” Teased and expected for months, the reasoniQuick Update on AI “Reasoning” Models as OpenAI Releases New o3 Variantsng model was known by the codename “Strawberry” since spring 2024, and a weaker version, “o1-preview,” had been accessible to paid ChatGPT Plus subscribers for a couple of months. Writing about “Strawberry/o1-preview” in September, Prof Ethan Mollick explains that these reasoning models solve complex problems by planning and iterating, excelling in science and logic tasks.

As an experiment, I asked o1 to craft a research plan. My PROMPT had several components: to provide background for the context window (the model’s “short-term memory” which functions like the setting of a painting), I requested the model review and summarize the GPS and best practices for creating a genealogical research plan; I then provided the model with some factual information from an ongoing project, excerpted from an old blog post.

Here is the full prompt. 

PROMPT:
1. Review and summarize the Genealogical Proof Standard. 
2. Review and summarize the best practices for crafting a genealogical research plan. 
3. Craft a genealogical research plan for the first case mentioned in this post excerpt:
<post>
I have two possible research objectives I am considering for my focus in this course, one modest, the other perhaps too ambitious. Both possible objectives are to confirm or refute the parentage of an ancestor using DNA analysis and documentary evidence. The more ambitious case dates to the late 1700s, involving my 3rd great-grandfather, Isaac Little (1799 — 1884, Ashe, NC) and determining Isaac’s father from among two brothers (Edmund Little or Peter Little) or perhaps their father, Charles Little (a 111-marker Y-DNA test strongly suggests that Isaac and this researcher are descendants of Charles’s ancestor, Abraham Little [1677, England — 1724, Virginia], mentioned above as my surname immigrant ancestor).
</post>

NOTE: My usual practice is to “chat with your chatbot,” that is, to iterate through a conversation with a language model to develop best response. But for this experiment, I wanted to see the model’s first, initial effort. You can judge for yourself below. And for comparison, you can evaluate o1’s response below to the relatively paltry response to the same prompt by the legacy model GPT-4.

When evaluating LLM responses, Prof Mollick suggests we ask, how does this response compare to: 1) the best human expert; 2) the average human worker; and 3) the average person; applying that rubric here, we might ask, how does this response compare to: 1) the best genealogist alive; 2) the average family historian; and 3) the average woman or man on the street.

The question I’d suggest asking, when evaluating this response, is: How long were you doing genealogy before you could write a research plan of this sort, given, say, 90 minutes (not just the 90 seconds it took o1).

ChatGPT was released in November 2022; in a manner of speaking, a two-year-old crafted this research plan. Can you imagine what a four-, seven-, or ten-year-old AI will be able to do? I can’t.

The reasoning model generated the response below. You can see the actual ChatGPT session here: https://chatgpt.com/share/67526ede-29fc-8004-87e7-227a5d70cfe6.

Genealogical Research Plan Generated by OpenAI’s First Full “Reasoning” Model, o1

Crafting a Genealogical Research Plan for Isaac Little (1799–1884, Ashe County, NC)

Research Objective:
Determine the paternity of Isaac Little (1799–1884) by confirming or refuting whether his father was Edmund Little, Peter Little, or possibly their father Charles Little. Utilize both documentary evidence and DNA analysis to strengthen or clarify the conclusion.

Background and Known Information:

  • Known Ancestor: Isaac Little (1799–1884), resided in Ashe County, North Carolina.
  • Timeframe: Late 1700s to early 1800s. Isaac’s birth around the turn of the century places the possible father’s generation in the late 1700s.
  • Potential Fathers: Brothers Edmund Little, Peter Little, or their father Charles Little.
  • DNA Context: A 111-marker Y-DNA test strongly suggests descent from Abraham Little (1677, England—1724, Virginia), an ancestor to Charles Little and presumably to his descendants. This narrows the paternal line to the Little family in question, but does not directly distinguish among Edmund, Peter, or Charles as Isaac’s father.

Step-by-Step Research Plan:

  1. Clarify Lineage and Construct a Preliminary Family Tree:
    • Assemble a family group chart for Abraham Little and his descendants, with special attention to Charles Little and his sons.
    • Document all known children of Charles Little (including Edmund and Peter), their dates, and residences.
    • Note any known siblings of Isaac Little and their records to see if patterns of association might emerge.
  2. Documentary Evidence Search:
    • Vital and Church Records:
    • Search for birth, baptism, or early church records in Ashe County and surrounding areas to identify any record referencing Isaac Little’s parents.
    • Check local church minute books or membership rolls circa 1790–1820 for mention of Charles, Edmund, or Peter Little and their families.
  3. Land and Property Records:
    • Examine Ashe County (or predecessor counties) deeds and land grants. Look for transactions involving Charles, Edmund, or Peter Little, and note any references to Isaac (for instance, if Isaac inherited land or served as a witness to a deed involving a likely father).
    • Consider whether Isaac occupied or sold family land, which may indicate inheritance patterns.
  4. Tax Lists and Early Census Records:
    • Review pre-1850 censuses and tax lists to identify household compositions of Charles, Edmund, and Peter Little. While pre-1850 censuses only list heads of households, patterns in the ages and sexes of household members might help locate Isaac as a minor in the correct household.
    • Identify continuous proximity of Isaac to either Edmund, Peter, or their heirs in later tax lists, which may suggest familial ties.
  5. Probate and Wills:
    • Investigate wills, estates, and probate files of Charles, Edmund, and Peter Little. If any left wills, did they name a son Isaac? If not named, can any estate distributions or guardianship records hint at an informal acknowledgment of Isaac as a son or ward?
    • Check guardianship records if Isaac was underage at the time of a father’s death.
  6. Court and Legal Records:
    • Look for lawsuits, disputes, or indenture bonds that might mention Isaac in relation to Edmund, Peter, or Charles.
    • Petitions, bastardy bonds (if that was a legal practice in that region), or other legal documents may provide parental clues if Isaac’s legitimacy or guardianship was ever at issue.
  7. Local Histories and Published Family Genealogies:
    • Review county histories, local biographical sketches, and published genealogies or family histories for the Little family. Although secondary, they may provide leads or references to original documents.
  8. DNA Analysis and Correlation:
    • Refine Y-DNA Matching:
      • Work with known Y-DNA matches who descend from Abraham Little lines. Identify matches known to descend from Edmund, Peter, or Charles’s other sons.
      • Compare genetic distance, shared marker mutations, and Time to Most Recent Common Ancestor (TMRCA) estimates among these match groups.
    • Autosomal DNA Testing (If Available):
      • Explore autosomal DNA segments shared with documented descendants of Edmund or Peter Little. Segment triangulation may help determine which branch of the family Isaac is more closely aligned with.
      • If possible, identify known living descendants of Edmund and Peter, and compare autosomal matches to see if Isaac’s descendants share more significant DNA segments with one branch over the other.
  9. Synthesis and Analysis:
    • Correlate the DNA evidence with documentary findings. For example, if probate records exclude Isaac from Edmund’s heirs but do not exclude him from Peter’s, and DNA evidence strongly points to Peter’s line, this combination strengthens that hypothesis.
    • Consider any conflicting evidence. If a deed suggests one paternity while DNA points another way, delve deeper into the records (perhaps the deed involved a different Isaac Little, or there were naming conventions that caused confusion).
  10. Conclusion Draft:
    • As information emerges, draft a written analysis showing why the available evidence supports one particular paternal identity over the others.
    • Cite all sources used and explain the reasoning process in accordance with the Genealogical Proof Standard.
    • If the conclusion remains tentative, identify what further records or DNA tests might resolve lingering doubts.
  11. Iterative Revisions:
    • Be prepared to adjust the plan as new findings surface. If a critical record type (like a surviving will or a set of church registers) comes to light, incorporate it and reassess earlier conclusions.

I wanted to share this with you as soon as possible, so I have not critiqued this response with a fine-toothed comb, but I’ll make note of a few quick observations:

  1. I was pleased to see that the model credited the Board of Certified Genealogists in the first sentence of its review of the Genealogical Proof Standard; earlier models frequently skipped attribution.
  2. The model seemed to do okay with organization and clarity, suggesting sources, integration of DNA and documentary evidence, and adherence to the GPS.
  3. The model could have done better with considering a wider geographical scope, prioritizing tasks (perhaps addressing probate and land records before tackling DNA segment triangulation), saying more about socio-historic contextual research, and addressing the limits of autosomal DNA at this generational distance.

Not too shabby for a two-year-old, though.

This was just Day One of the Twelve Days of OpenAI, so there will be more to come this holiday season.

Episode #19 of The Family History AI Show podcast is out now; in that episode, my co-host, Mark Thompson, and I discuss the top five AI breakthroughs for genealogists in 2024. And we just recorded Episode #20 a few hours ago and we talk about reasoning models and the release of full o1, so look for that around the middle of next week.

If something big happens during these Twelve Days of OpenAI, we may do our first “emergency” podcast 😉. I don’t expect Orion/GPT-5 to drop this month, but if something like GPT-4.5, Sora (video generation), improved image generation and analysis (HTR!), and/or an OpenAI web browser were to be released, that would be a nice holiday treat.

Exploring Facebook’s New AI Tool for Groups

I recently stumbled upon a new AI feature in Facebook that caught my attention. It all started when I saw an intriguing image from a recommended group—Certified Arborists—and clicked through to check out more. What greeted me wasn’t just a series of posts but a new AI assistant for the group, something I hadn’t expected. This AI, dubbed “ArborMate,” seems to be part of Facebook’s larger push into artificial intelligence, an effort that doesn’t get as much buzz as Elon Musk’s Grok for X, but is equally ambitious.

Meta, Facebook’s parent company, has been quietly but steadily investing in AI, and their goal is clear: by 2025, Meta AI intends to be on the same footing as OpenAI, Anthropic, and Google. With deep resources, massive data stores, and the technical leadership to match, they have the potential to be a serious contender in the AI space. These group AI features are likely just the beginning, and while not everyone will embrace them, their potential to provide useful, educational experiences is hard to ignore.

Image 1: Meet ArborMate, the group’s custom AI!

Testing ArborMate: A Look at Group AI in Action

The Certified Arborists group AI, named ArborMate, is particularly interesting for a few reasons. First off, even as a non-member, I was able to use the AI to interact with the group discussion data. The group is set to Public, meaning there was no expectation of privacy—and that allowed me to test the bot’s capabilities freely.

Image 2: The Certified Arborists group on Facebook features an AI-enabled assistant.

My first prompt was my typical opener whenever I encounter a new chatbot: “Who are you? What do you know? What can you do?” The AI assistant’s response was polished, on par with what you might expect from other major models like ChatGPT or Google’s Gemini. It claimed to be capable of answering questions about arboriculture based on the group’s discussion data, and it gave a pretty comprehensive overview of its capabilities.

Image 3: ArborMate, the AI assistant for the Certified Arborists Facebook group, offers to answer tree-related questions based on the group’s discussions, providing relevant and accurate information.

Diving Deeper: Search and Summarization

What impressed me next was the AI’s ability to perform a deep search of the group’s historical discussions. For example, I asked it to “Search the discussions for mentions of Cherry trees.” The response was crafted from various earlier posts within the group—and more impressively, each statement was linked back to its source. This level of transparency and traceability really highlights how an AI like ArborMate can be used as an effective tool for finding specific information quickly, a significant improvement over the usual keyword searches.

Image 4: ArborMate responds to a query about cherry trees, summarizing several past discussions.

I then pushed the assistant a bit further. My next prompt was: “What are folks in the Certified Arborists group most concerned about?” The AI summarized the key issues: tree health, safety, and professional services. It seemed to pull in the wisdom of the group, synthesizing many perspectives into one coherent answer.

Image 5: ArborMate summarizes the key concerns of the Certified Arborists Facebook group.

A Structured Test: A Chain-of-Thought Prompt

Finally, I ran a structured prompt to see how well it could generate a plan based on collective insights. My prompt was:

“Tell me all about: Removing hazardous trees or branches. And do it this way: A) State your initial assessment and plan.B) Review and critique your plan.C) Revise and improve your plan.D) Execute your plan.”

The AI compiled a surprisingly detailed answer, again citing sources throughout, and showed an understanding of risk, safety, and proper techniques—essentially giving a nuanced and phased approach to the task. For a group like this, having access to a synthesized body of knowledge could be invaluable.

Image 6: ArborMate provides a structured plan for removing hazardous trees.

Closing Thoughts: The Potential and Pitfalls

I am not a certified arborist, and probably you aren’t either, and I hope we have few tree emergencies. But think about all the Facebook groups you belong to. There are definitely some groups where having this kind of AI access could be a gold mine—like for local genealogical societies with active discussions. For other groups, it might not be appropriate, and I hope group owners will have the option to decide that for themselves.

In the meantime, the experiment with ArborMate shows a glimpse of where social media is heading—a place where not just human, but also AI-driven conversations, can enrich our shared knowledge.

Suggestions for Facebook Group Admins

If you’re a group admin considering whether to enable an AI assistant like ArborMate, here are some tips to help you decide:

  • Review Privacy Settings: Ensure your group’s privacy settings align with how you want the AI to function. Public groups may allow broader access, including to non-members, so understand the implications of your current settings.
  • Opt-In Considerations: Be aware that AI features may become available on an opt-in basis. Make sure you fully understand what the AI will do and how it will use group data before enabling it.
  • Assess Group Relevance: Think about whether your group’s content would benefit from AI-assisted summaries, search, and discussion analysis. Groups focused on technical or educational topics may find AI especially useful.
  • Communicate with Members: Inform your group members about the AI and how it will be used. Transparency will help build trust and ensure members are comfortable with its role in the group.
  • Moderate AI Use: Monitor how the AI is being used and be ready to make adjustments. There may be situations where the AI needs fine-tuning, or where its presence isn’t adding value.

These suggestions can help ensure that the AI becomes a helpful resource while respecting your group’s culture and member expectations.

Note: This post was originally written organically as a Facebook post at Blaine Bettinger’s group “Genealogy and Artificial Intelligence (AI),” the center of gravity for discovery, discussion, and learning about AI-assisted genealogy; I used OpenAI’s Canvas to re-work my original posts and comments there into this post.

Paper-to-Podcast Prompt

In this post you’ll find:
● a PROMPT to generate a podcast script/transcript from any type of content
● an example of a podcast script/transcript generated by this PROMPT using a prompt engineering resource as a source document, i.e., hear two hosts discuss prompt engineering in an easy, accessible style
● a list of upcoming speaking events

A new AI tool has been generating a great deal of interest the past few weeks. Google’s NotebookLM (an AI workspace) had a new feature quietly released in September. (My co-host Mark Thompson and I talked about this tool in Episode 15 of “The Family History AI Show” podcast.) Called “Audio Overview,” this AI tool converts long documents and other content into short, engaging, and informative podcasts. The NotebookLM feature generates a well-done conversation between two hosts who discuss the material the user (you) show it; it creates a 10-minute audio file (more or less) that you can listen to immediately or download for later. Folks have been tremendously impressed. I was. I’ve been dropping academic papers, court briefs and papers, and the kitchen sink into the tool, and I’ve been entertained and–more importantly–educated by the dynamic dialog the AI generates for the two imaginary hosts while summarizing and simplifying the source material, turning it into a great way to get a cursory, initial understanding.

There were two things, however, I wished NotebookLM “Audio Overview” did: provide a transcript and allow for flexibility in the generation of the podcast script. So, I fixed it. Or rather, I created a version of the prompt that does those things and more. Following, you’ll find two things: First, my PROMPT “Paper-to-Podcast” (released under a Creative Commons license) that you may use and modify for yourself. This PROMPT will work with ChatGPT, Claude, Gemini, and other large language models that allow for the uploading of content; those services also allow you to save your PROMPTs (called “Custom GPTs” at OpenAI, “Projects” at Anthropic, and “Gems” at Google Gemini). Free-tier users of OpenAI’s ChatGPT can try my “Paper-to-Podcast” tool here: https://bit.ly/Paper-to-Podcast. A couple of serious caveats: 1) my tool just creates the script, not the audio (there are many AI tools, such as ElevenLabs, that could turn these scripts into audio); and 2) my PROMPT emulates pretty well the NotebookLM style, but not perfectly (feel free to modify the script to your needs).

The “Paper-to-Podcast” Prompt

Copy-and-paste the PROMPT below into your favorite chatbot or use it to create your own Custom GPT, Project, or Gem (i.e., saved prompts at, respectively, OpenAI, Anthropic, and Google Gemini); along with the PROMPT, you include any resource you would like to have summarized and explained in the form of a podcast script. Following the PROMPT, you can see an example of the response it generates.

<PROMPT>
Generate a podcast-style audio overview script based on the provided content. The output should be a conversational script between two AI hosts discussing the main points, insights, and implications of the input material.

Podcast Format:
- Duration: Aim for a 10-minute discussion (approximately 1500-2000 words)
- Style: Informative yet casual, resembling a professional podcast
- Target Listener: A busy professional interested in efficient information consumption and staying updated on the latest developments in the field

Host Personas:
- Host 1: The "Explainer" - Knowledgeable, articulate, and adept at breaking down complex concepts
- Host 2: The "Questioner" - Curious, insightful, and skilled at asking thought-provoking questions
- Relationship: Collegial and respectful, with a hint of friendly banter

Podcast Structure:
1. Introduction (30 seconds; 100 words):
   - Briefly introduce the hosts and the topic
   - Provide a hook to capture the listener's interest

2. Overview (1 minute; 200 words):
   - Summarize the key points from the input content
   - Set the stage for the detailed discussion

3. Main Discussion (7-8 minutes; 1600 words):
   - Analyze and discuss the most important aspects of the topic
   - Present different perspectives and potential implications
   - Use specific examples and details from the input content to illustrate points

4. Conclusion (30 seconds; 100 words):
   - Recap the main takeaways
   - Provide a thought-provoking final comment or question

Content Analysis and Discussion:
- Identify the core concepts, key arguments, and significant details from the input material
- Organize the discussion around these main points, ensuring a logical flow of ideas
- Encourage a balanced exploration of the topic, considering various viewpoints when appropriate

Tone and Style:
- Maintain a conversational, engaging tone throughout the discussion
- Use clear, accessible language while accurately conveying complex ideas
- Incorporate natural speech patterns, including occasional "disfluencies" (e.g., "um," "uh," brief pauses) and conversational fillers (e.g., "you know," "I mean")
- Add moments of light banter or personal observations to enhance the natural feel of the conversation

Handling Sensitive Topics:
- Approach potentially controversial subjects with neutrality and objectivity
- Present multiple perspectives without showing bias
- Use phrases like "Some argue that..." or "Another viewpoint suggests..." to introduce different opinions

Script Refinement Process:
1. Generate an initial outline of the discussion
2. Develop a detailed script based on the outline
3. Review the script for clarity, coherence, and engagement
4. Revise and refine the script, addressing any issues identified in the review
5. Add natural speech elements, banter, and "disfluencies" to the polished script

Additional Guidelines:
- Seamlessly incorporate specific examples, quotes, or data points from the input content to support the discussion
- Ensure that the hosts complement each other, with the "Explainer" providing in-depth information and the "Questioner" driving the conversation forward with insightful queries
- Maintain a balance between informative content and engaging dialogue
- End the podcast with a statement or question that encourages further thought or discussion on the topic

Remember to generate a script that sounds natural and engaging when read aloud, as if it were a real-time conversation between two knowledgeable hosts.

<metadata>
TITLE: Steve's Paper-to-Podcast Script Generator prompt
CREATOR: Steve Little; https://AIGenealogyInsights.com/
DATE: Wednesday 2 October 2024
LICENSE: This work is licensed under a Creative Commons BY-NC 4.0 License.
INSPIRATION: Inspired by Google's NotebookLM "Audio Overview"
</metadata>

</PROMPT>

This “Paper-to-Podcast” prompt is designed to transform dense, complex content into engaging and accessible podcast scripts. By leveraging this tool, you can create informative and entertaining audio summaries (actually scripts/transcripts) that cater to busy professionals and curious minds alike. Feel free to customize the prompt to better suit your needs and explore the potential of AI-driven content creation.

An Example: A Podcast about Prompt Engineering

There are three ingredients to generate these scripts: the PROMPT (above), a source (a document, image, text file, PDF, etc.) to talk about, and an LLM to generate the response. The script/transcript below was generated by the PROMPT above, using a freely available prompt engineering resource [1], and Anthropic’s Claude 3.5 Sonnet. Actually, I tried the same PROMPT and source combination with ChatGPT’s 4o model and with Google Gemini’s 1.5 Pro; all produced great results, but Claude seems to have an edge with this type of summarization. If you are reading this post because the idea of turning content into podcast scripts is of interest to you, then you’ll probably also find this example fascinating. I promise that Mark and I don’t use this to generate our show. But perhaps we oughta.

The AI Insider (fictional podcast): Demystifying Prompt Engineering

Sarah (Explainer): Welcome to The AI Insider, your weekly dose of cutting-edge artificial intelligence insights. I'm Dr. Sarah McQueen.

Alex (Questioner): And I'm Alex Johnson. Today, we're diving into a topic that's been buzzing in the AI community: prompt engineering. Sarah, what exactly is prompt engineering, and why is it suddenly so important?

Sarah: Great question, Alex. Prompt engineering is essentially the art and science of crafting instructions for large language models, or LLMs, to get them to produce the output we want. It's become crucial as these models are increasingly integrated into various applications.

Alex: Interesting! So, it's like learning how to "speak AI," right? How does this process actually work?

Sarah: That's a great way to put it, Alex. At its core, prompt engineering involves four key components: instructions, context, input data, and an output indicator. Let's break these down.

Alex: Alright, walk us through it, Sarah.

Sarah: Sure thing. First, you have the instructions, which are clear directives telling the model what to do. Then there's context, providing relevant background information. Next, you have the input data, which is the actual content the model will work with. Finally, there's the output indicator, specifying what kind of response you're looking for.

Alex: Hmm, can you give us an example to illustrate this?

Sarah: Absolutely. Let's say we want to classify a restaurant review. Our prompt might look like this: "Classify the following review as positive, neutral, or negative: 'The food was delicious, but the service was slow.' Sentiment:"

Alex: Oh, I see. So the instruction is to classify the review, the context is that we're looking at restaurant feedback, the input data is the actual review, and the output indicator is the word "Sentiment:" at the end.

Sarah: Exactly! You've got it, Alex.

Alex: Now, I've heard that there are some settings you can adjust when working with these models. What can you tell us about that?

Sarah: You're right, there are two key parameters to consider: temperature and top-p. Temperature controls how random or creative the model's responses are. A lower temperature produces more focused, deterministic answers, while a higher temperature encourages more diverse and creative responses.

Alex: Fascinating! And what about top-p?

Sarah: Top-p, also known as nucleus sampling, narrows down the range of possible word choices. It helps control the breadth of the generated text, allowing for more or less variability in the output.

Alex: So, if I'm understanding correctly, you'd use different settings depending on whether you want a more factual or creative response?

Sarah: Precisely! For instance, if you're working on a creative writing task, you might set a higher temperature, like 0.7. But for fact-based responses, you'd want a lower temperature, maybe around 0.2.

Alex: That makes sense. Now, Sarah, I'm curious about the different tasks prompt engineering can be applied to. Can you give us some examples?

Sarah: Of course! Prompt engineering is incredibly versatile. It can be used for text summarization, question answering, text classification, and even code generation, among other tasks.

Alex: Wow, that's quite a range! Could you elaborate on one of those?

Sarah: Sure, let's take text summarization as an example. Say you have a lengthy medical article, and you want to get the key points quickly. You could craft a prompt asking the model to summarize the main ideas in one sentence. This could be incredibly useful for busy professionals who need to stay updated but don't have time to read entire articles.

Alex: That sounds really practical. Are there any advanced techniques in prompt engineering that our listeners should know about?

Sarah: Absolutely, Alex. Three techniques that have shown promising results are few-shot prompting, chain-of-thought prompting, and self-consistency.

Alex: Those sound intriguing. Can you break them down for us?

Sarah: Certainly. Few-shot prompting involves providing a few examples in the prompt to guide the model. It's like giving the AI some context to work with. Chain-of-thought prompting encourages the model to explain its reasoning step by step, which is particularly useful for complex problem-solving tasks.

Alex: And what about self-consistency?

Sarah: Self-consistency is a technique where the model generates multiple reasoning paths and then selects the most consistent answer. It's particularly effective for tasks involving common-sense reasoning.

Alex: These techniques sound powerful. Are there any tools available to help with prompt engineering?

Sarah: Yes, there are several libraries and integrated development environments (IDEs) that support prompt engineering. These tools allow users to experiment with different prompts, evaluate their effectiveness, and even manage prompt versions.

Alex: That's really helpful. Now, looking towards the future, what trends do you see emerging in prompt engineering?

Sarah: Great question, Alex. One exciting development is Reinforcement Learning from Human Feedback, or RLHF. This technique aims to align model responses more closely with human preferences. We're also seeing a growing focus on multimodal prompting, which involves working with models that can handle not just text, but also images and video.

Alex: Fascinating stuff, Sarah. But I have to ask - are there any concerns we should be aware of?

Sarah: Absolutely, Alex. As with any powerful technology, there are safety concerns. Two significant issues are prompt injection and jailbreaking, which can potentially manipulate model outputs or bypass safety protocols. These are active areas of research in the AI security community.

Alex: It sounds like prompt engineering is a complex and rapidly evolving field. Any final thoughts for our listeners?

Sarah: I'd say that prompt engineering is becoming an essential skill in the AI toolkit. It's not just about getting better results from models; it's about understanding how to communicate effectively with AI systems. As these models become more integrated into our daily lives and work, the ability to craft effective prompts will be invaluable.

Alex: Well, there you have it, folks. A deep dive into the world of prompt engineering. Thanks for tuning in to The AI Insider. I'm Alex Johnson.

Sarah: And I'm Dr. Sarah McQueen. Until next time, keep exploring the fascinating world of AI!

If you have a question you’d like to ask, or you made a helpful modification to the PROMPT that you’d like to share, or follow along with the discussion, please look for a thread about this post in the “Genealogy and Artificial Intelligence (AI)” group at Facebook. Thanks, Steve

Appearances

  • Co-host Mark Thompson and I will be Dana Leed’s guest as she livestreams from Baltimore on Friday 4 October 2024 for East Coast Genetic Genealogy Conference (ECGGC); catch her YouTube channel at 11 AM ET for about an hour of AI genealogy discussion and Q&A.
  • I will be speaking at ECGGC at 7 PM ET, Friday 4 October 2024 with Mark Thompson; this talk will be livestreamed for Conference registrants.
  • I will be livestreaming a presentation for the Eastern Washington Genealogical Society at 1 PM ET (10 AM PT), Saturday 5 October 2024, as part of their “Future of Genealogy” seminar.

Footnotes:

Episode 12: Hollywood AI Blunder, AI Image Generator Roundup, Google Lens Saves You Time Researching, Use AI For Translation

In this week’s episode of “The Family History AI Show,” my co-host Mark Thompson and I take you on an exhilarating journey through the latest AI advancements revolutionizing genealogy. We kick things off with a cautionary tale from Hollywood, highlighting the critical importance of fact-checking AI-generated content. Then, we dive deep into the world of AI image generators, exploring tools like DALL-E, Midjourney, and Adobe Firefly, and discussing their potential to breathe new life into family history narratives. We’re particularly excited to share our insights on the game-changing updates to Google Lens and its integration with Chrome, which we believe will transform how genealogists interact with online content. Our “Tip of the Week” focuses on the versatile applications of AI in translation, going far beyond simple language conversion. We wrap up with our “RapidFire” segment, covering everything from Microsoft’s new Tab Organizer to the future of AI agents in research. Whether you’re a tech enthusiast or just starting to explore AI’s role in genealogy, this episode is packed with practical advice and inspiring ideas to enhance your family history research. Join us for an hour of engaging discussion that will expand your horizons and revolutionize your approach to uncovering your family’s past.

https://blubrry.com/3738800/136034321/ep12-hollywood-ai-blunder-ai-image-generator-roundup-google-lens-saves-you-time-researching-use-ai-for-translation

Episode 11: Microsoft Simplifies Research with “Deep Search”, Google Goes All In on AI, Have We Entered the Post-Photography Era?

In this episode of “The Family History AI Show” podcast, hosts Mark Thompson and I discuss Microsoft’s new AI-enhanced search tool, Deep Search. Then, we review Google’s major list of AI announcements. In the last of the big stories this week, Mark and I talk about how photography is changing in the digital age.

Don’t miss this week’s Tip of the Week, where Mark provides valuable insights on crafting better prompts using the role-goal-task-format style.

The rapid-fire segment covers recent developments in AI image generation, including updates from OpenAI’s DALL-E and Midjourney. We also examine OpenAI’s safety measures and Elon Musk’s controversial AI tools.

Blending expert analysis and practical advice, this podcast equips listeners to navigate the exciting world of AI in genealogical research and beyond.

https://blubrry.com/3738800/135489123/ep-11-microsoft-simplifies-research-with-deep-search-google-goes-all-in-on-ai-have-we-entered-the-post-photography-era