2025: What's Ahead?

The Wright Brothers test fly their aircraft on Fort Myer's parade field

2025: What's Ahead?

This year in genealogy research has been momentous. Just as the Wright brothers' first flight revolutionized travel, 2024's genealogical advances are transforming how we explore our family histories.

The Wright Brothers' Legacy

On 17 December 1903, the Wright brothers, Orville and Wilbur, made their first successful sustained flight. Some of the developments we've seen this year will change the amount of information available to us much like that event changed history, transportation, war, travel, and the way we humans viewed the world. In addition, we are getting tools to help us access, analyze and use that information to tell the stories of our families so they will not be lost.

Before that fateful flight, we humans did not experience the entire world as being within our reach. Most of the world had rail, but to cross the oceans, one traveled to a port city and got on a ship. By June 1919, only 16 years after that first sustained flight, British aviators had crossed the Atlantic by airplane. Now many of us have connected with friends and relatives all over the world, both virtually and in international travel.

Just as that first flight helped catapult us to the moon and beyond, the field of computer-assisted tasks of all sorts are growing at a dizzying pace. Machine learning and the development of AI is growing ever more quickly now. When I asked Perplexity.ai to "give me a brief history of machine learning and AI:

Perplexity: "Key Developments

1960s-1970s:

  • Development of expert systems
  • Creation of ELIZA, the first chatbot 
  • Introduction of backpropagation for neural networks

1980s-1990s:

  • Shift from knowledge-driven to data-driven approaches
  • IBM's Deep Blue defeated world chess champion (1997) 

2000s-Present:

  • Emergence of deep learning (2006) 
  • Advancements in natural language processing and computer vision
  • Development of intelligent agents and virtual assistants
Machine learning and AI continue to evolve rapidly, with recent breakthroughs in areas such as generative AI and large language models."

Source links 1. https://www.techtarget.com/whatis/feature/History-and-evolution-of-machine-learning-A-timeline, 2. https://www.algotive.ai/blog/the-history-of-artificial-intelligence-machine-learning-and-deep-learning


The improvements announced this past year have changed genealogy research forever; and we researchers were already being helped in ways many of us didn't realize were machine learning or AI work products. Consider the amazing speed of the creation of 1950 Census indexes after it was released to the public April 1, 2022. When asked "what software was used to index the US 1950 Census?" Perplexity replied: 

"To create the initial name index for the 1950 Census, NARA used a combination of technologies: Optical Character Recognition (OCR) and Artificial Intelligence (AI) and Machine Learning (ML)." They were not using the newest tech, but it worked. quickly!

 Ancestry developed their own index and shared it with FamilySearch, who provided the humans who checked all the computer-generated text. Organizations and researchers collaborated with the improving AI tech. Fortunately, humans were "in the loop" which ensured both better data for all of us, and also improved the software for the next huge task. We now know that was the FamilySearch Full Text Search. Click the link for direct access to 3455 collections as of 19 December 2024. Each time this new tool is brought up in conversation, there are multiple stories of long-standing brick walls dissolved. 

What genealogical breakthroughs have you experienced this year? Share your stories in the comments!


I can't wait for Rootstech 2025 to hear what's next!


~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~


Send your stories to m.strickland@skcgs.org


Comment


Subscribe


Valorie Zimmerman

🌞 My thanks to Perplexity, Bob Zimmerman and MaryLynn Strickland for their help and feedback on this post.

Comments

  1. While in Facebook to link to this post on the SKCGS FB page, I read a PDF given out in the Genealogy and Artificial Intelligence (AI) group. It said: 22 Nov 2024 news quote from the National Archives Catalog
    that demonstrates the real potential of the LLMs for genealogical purposes:
    “We are excited to announce that through a partnership with FamilySearch, AI extracted text for all
    2,322,137 pages of the Case Files of Pension and Bounty-Land Warrant Applications Based on
    Revolutionary War Service, ca. 1800 - ca. 1912 is now available in the National Archives Catalog.
    “FamilySearch used 30,000 pages of pensions transcribed by our Citizen Archivists to teach their AI
    language model how to transcribe this 19th Century cursive writing. The Partner Contributed
    transcriptions generated by their AI language model for the entire series will now be available as
    Extracted Text in the National Archives Catalog.” The link to the database: https://catalog.archives.gov/id/300022?utm_campaign=AIandRevWarPensions&utm_medium=email&utm_source=newsletter

    ReplyDelete

Post a Comment

Popular posts from this blog

Looking for a Needle in a Haystack? GAME CHANGER at FamilySearch

What's the Question?

Help Save This Abused and Neglected Cemetery