Something I learned From Chatting with AI

I have been working with tax records in Beaver County for my husband’s Gorrell family. The tax records for the family range from 1804 through 1840 across two or three different townships. Some years there are one or two entries, other years three or four. I recorded the entries by year and included the census entries from 1800 to 1840. These are census records with only the head of household and tick marks for the other members by age and sex. I also included some of my own comments and source citations. I ended up with 9 pages. The work I’m doing is to find the family of Amos Gorrell (1804-1890).

Here is an example of what I created:

1800 South Beaver Township census. The only townships at this time were First Moon, Hanover, North Beaver, Second Moon, Sewickley, and South Beaver. The Gorrells were enumerated in South Beaver Township. Two men were enumerated next to each other. One James is older than the other, also indicated by the senior and junior appendages to their name. It is possible the James senior is James junior’s father. It is also possible one is just older than the other and not related as father and son.

James Gurl Senr. 1 male over 45, 1 female over 45.
James Gurl Junr. 4 males under 10, 1 male 26-44, 2 females under 10, 1 female 26-44.[1]

COMMENT: If James Senr was over 45, he was born before 1755. James Junr was between 26 and 44, so born between 1756 and 1774. If he was born closer to 1774, then James Senr could be his father. James Junr had six children all under 10, so the first could have been born about 1790, which may push his age closer to 1770, the proposed age for Amos’ father.

1804, South Beaver.
Thomas Grirel, 1 horse and 1 cow [not sure if this person was a Gorrell]
James Gorrel Sen, 2 horses[2]
James Gorrel Jun or James Dunkin, 200 acres, $100
James Gorrel Jun, 200 acres ($100), 200 acres ($50), 2 horses, 2 cows[3]

COMMENT: The James Gorrell Junior has land. No deed has been found for the purchase of this land in Beaver County, suggesting that the land may have been purchased in Alleghany or Washington County before the formation of Beaver County in 1800. No deed for the land has been found in Alleghany or Washington County. I am not sure how James Dunkin (or Duncan) fits in and I’m not sure why James Junior was listed twice. It is possible that either James Dunkin was renting from James Gorrell or vice-versa.

1805, South Beaver.
James Gorrell Junr, 200 acres, 1 horse, 2 cows.[4]
James Gorrell Senr, 1 horse, 1 cow.[5]

1806, South Beaver.
James Gorrel Senr, $.17
James Gorrel, Junr, $1.13.[6]

1807, South Beaver.
Elizabeth Gorrell, 1 horse, 1 cow.
James Gorrell, 300 acres, 1 horse, 2 cows.[7]

COMMENT: Between 1804 and 1806, there were two men named James Gorrell taxed. One was listed as senior and the other junior. There is no guarantee that they are father and son, but could be. The elder Gorrell did not own land, only a horse and cow. The younger one owned land. In 1807, Elizabeth Gorrell is taxed for the horse and cow, making her likely the widow of James Gorrell Senior, who may have died between 1806 and 1807. The James listed here is likely the James junior from earlier tax records.

Work with AI
I continued in the same manner until I got to the end of the available tax records in 1841. I wasn’t sure where to go from here. It was a lot of data. It needed to be organized in a different way. I thought AI could help. So, I gave Claude Opus 4.8 high this prompt and the document:

I have created a research report about the Gorrell family in Beaver County, Pennsylvania where I found various family members in tax and census records from 1800 to 1840. The family was found in different records by different spellings of the last name. These families were found in Ohio and South Beaver townships. I will upload the Word document. What conclusions can be discovered from the data? Organize the data in another way so I see each person's records separately. What other resources should be searched next?

It took a long time for the analysis and then I received a 21-page report explaining how it organized the work, along with explanations and analysis. It identified 14 distinct people along with two incidences of Gorrell heirs. It then created tables of each identity with the same data that I had collected. This would have taken me a long time to reorganize the data. Here is an example of the first identity. I should have had it include the source citations I provided.

Year	Twp.	Name as written	Land	Livestock	Reading
1800	S.Bvr	James Gurl Senr (census)	—	—	1 male 45+, 1 female 45+. Born before 1755. No children at home.
1804	S.Bvr	James Gorrel Sen	none	2 horses	Taxed on movables only. No real estate.
1805	S.Bvr	James Gorrell Senr	none	1 horse, 1 cow	Herd reduced by one horse.
1806	S.Bvr	James Gorrel Senr	—	—	Tax of $0.17. Consistent with a landless assessment.
1807	S.Bvr	— absent —	—	—	Elizabeth appears in his place with the identical 1 horse, 1 cow.

After each identity, it also had its analysis of the information. The mistake I made with the prompt was asking for it to analyze the data before I had a chance to see it reorganized. It certainly gave me a lot of possible conclusions that I did not have a chance to conclude myself. Would I have made the same conclusions about the data?  I feel very regretful.

Conclusion
My advice in using AI with genealogy data is to first have it organize the data into other forms. Then study the data yourself, recording your observations. If you think you may have missed something, then have the AI make its own observations about the data. This way, you can be sure that you have done your own work first.

I will admit that I am not very experienced in using AI. I dabble here and there. I am a bit uncomfortable about the energy and water usage of the large data centers. I believe genealogists should think for themselves. The best use of AI is the reorganization of the data. It does the work quicker. You still must double-check its work, but I never asked it to find information for me. It used the data I gave it. I also used Incognito mode, so the work we did together wasn’t training the model.

#52Ancestors: Week 30 – Free Space

This is my ninth year working on this year-long prompt, hosted by Amy Johnson Crow (https://www.amyjohnsoncrow.com/) at Generations Cafe. I write each week in one of my two blogs, either Mam-ma’s Southern Family or My Trails into the Past. I have enjoyed writing about my children’s ancestors in new and exciting ways.

---

[1] 1800 U.S. census, Beaver Co, Pennsylvania, South Beaver Twp, p. 169 (penned), lines 1 & 2, James Gurl Senr & James Gurl Junr.

[2] Beaver Co, Pennsylvania, tax record, 1804, South Beaver twp, p. 10, Thomas Grirel & James Gorrel Sen.

[3] Beaver Co, Pennsylvania, tax record, 1804, South Beaver twp, p. 11, James Gorrel Jun, James Dunkin, and James Gorrell Junr.

[4] Beaver Co, Pennsylvania, tax record, 1805, South Beaver twp, p. 12, James Gorrell Junr.

[5] Beaver Co, Pennsylvania, tax record, 1805, South Beaver twp, p. 11, James Gorrell Senr.

[6] Beaver Co, Pennsylvania, tax record, 1806, South Beaver twp, p. 6, James Gorrel Senr & James Gorrel Junr.

[7] Beaver Co, Pennsylvania, tax record, 1807, South Beaver twp, p. 7, Elizabeth Gorrell, James Gorrell.

Copyright © 2011-2026 by Lisa S. Gorrell, My Trails into the Past. All Rights Reserved.