The Future of Genealogy?
Dark Matter Structure Courtesy https://newatlas.com/dark-matter-filaments-found/23281/ |
Physics
Scientists estimate that there is five times the amount of "dark matter" in the universe than regular matter. Matter is everything that we can see, measure, or experience outside of thought, imagination and spiritual experience. But dark matter can so far only be simulated, although it must exist because of the powerful gravitational force it exerts within our cosmos.The latest NOVA on PBS, "Decoding the Universe: Cosmos" is about the discoveries in astrophysics in the past 50 years. An image of the theoretical structure of this dark matter struck me so strongly I had to pause the program and begin writing to you and thinking about why that image stopped me. The scientists are working with dark matter, and dark energy, about which they still know little, and yet it surrounds us, unseen.
Re-watching this NOVA, I'm struck over and over by the parallels between our research and that work done by the physicists over the past fifty years. Again and again, they have re-checked the data, gathering new, more precise data and checked old theories about how things work. Sometimes previous errors have been corrected, and just as in our research, new findings lead to more questions. For us, a fresh look often leads to records not previously uncovered, and as above, leaves us with more questions. For instance, DNA evidence is taking many of us up lines of our family trees which were completely undiscovered before.
We work with the unknown as genealogists and family historians, uncovering relationships and family networks which connect not only locally in the past, but even reach beyond the oceans, and through time, sometimes to the present.
Earlier, I was able to re-watch a great panel discussion about AI from the NGS National Conference. Blaine Bettinger had asked each panelist to imagine their answer to a question five years hence; and as I heard Dana Leeds' response, could only think of a supernova.
Supernova Remnant W49B; original from NASA. Digitally enhanced by rawpixel. Public Domain. |
Dana Leeds spoke about FamilySearch's Full Text search expanded by even more AI, to allow us to focus not just on single records, but for example draw maps of the land ownership of each county through time and watch a dynamic map show how the ownership changes. More than that, other early records such as tax lists, census and marriage records could also be brought in to enrich the data over time. What could be studied in such a way? Not just families, but both kin and FAN networks which surround us. The conception of such dynamic systems matched the NOVA images, and paralleled the description of dark matter and dark energy.
NGS
Blaine Bettinger: "We're 5 years from now at the 2029 NGS Conference, and each of us is presenting about AI. What are one or two things you are presenting or using the most, and how have they changed genealogy?"
Steve Little talked about "agents" which are tools that "take your Swiss Army knife and whittle whatever [tool] you need."
Mark Thompson: "All of the genealogy websites, my DNA, my family tree, all the local raw research notes, all of the family archive, all my photos...it's all accessible to my chat bot securely, privately, just for me, and I can ask natural language questions against it."
Dana Leeds: I'm going back to FamilySearch Labs and how they are doing every word search of the wills and deeds. In five years, there is just massive amounts of data, but we can talk about, instead of researching our little family and the deeds that they were part of, we can get the deeds for an entire county and we can follow that over time. And we can ask, in 1903, who was owning what land?...Or we can follow tax lists, and as we see people coming in and off a tax list, we can figure out that they must have migrated together. Massive amounts of that fan club, the friends, associates and neighbors, and just tying that in together."
- "Keynote Panel on Artificial Intelligence and Genealogy: A Discussion of Issues and Concerns" at NGS Family History Conference, 18 May 2024: Blaine Bettinger, Dana Leeds, Steve Little, Mark Thompson.
Later...
Steven Johnson, interviewed by Walter Isaacson about his book, "The Infernal Machine", says about J. Edgar Hoover and his role in creating scientific policing, "when historians catalog the momentous inventions in history--the printing press, the telescope, the steam engine--they rarely include indexing algorithms in their canon of breakthrough ideas. But tools that help us explore ever larger pools of information--and widen the net we can cast in those pools--have often turned out to trigger inflection points in history." J. Edgar Hoover was trained as a librarian, worked in the Library of Congress, and learned their new indexing system, described here: Library of Congress Classification. He adapted that to the Bureau of Investigation which became the F.B.I.; his system developed into what we now call a relational database.
The surprise at the end of the interview was what Johnson is now doing, which is working for Google, helping to create "Notebook LM" which is a LLM (large language model) chat bot that trains on your own data. (Amanpour and Company, Season 6, episode 322). It felt like Mark Thompson's prediction above is already happening!
Comments
Post a Comment