If you believe AI, it will tell you we just discovered the Treasure of the Sierra Madre
Or the Downfall of the Human Intellect
An Essay, Fred Holabird, 2024
We received a phenomenal batch of letters from the Coelln Collection from Arizona Territory (1883-1886) that were so good, I spent the better part of a week researching the man who wrote them and the circumstances around his detailed discussions of events.
Then somebody suggested I try a brand new AI research tool … uh oh, you won’t want to miss the factual comments below …
Robert Matheson, a Scotchman, transplanted from the San Jose, California area to Bisbee, Tombstone and the Sonoran/Sierra Madre area of Mexico in 1882 in pursuit of finding a successful mine. Like other mine engineers, his first jobs there were as a mine foreman for existing producing companies, though these were all small. Matheson was quick to make friends, including one of the local Bisbee assayers, John C. Tappenir. While Matheson appears to have arrived early in 1882, there is no mention of the Earps or the related incidents after the OK Corral in any form.
The essays on these letters will be published sometime before our April 6 auction, which are sure to fascinate anyone interested in western history. But for now, the short version…
Matheson prospected in southern Arizona and northern Mexico in the Sierra Madre, also known as, in general, “Sonora.” He located a rich gold area, hired seven miners, and went to work. His mine was attacked by Apache, killing all but Matheson. The details are gruesome. He escaped, but barely, and by the time he got back to town, the rumor had spread that the Apache had killed everyone. In essence, they threw a party for Matheson, the sole survivor. He soon arranged a posse to “avenge” the killings. Coincidently, General Cooke came into town three days later on a mission to quell the Apache uprisings, and Matheson’s posse joined up with Crooke and his huge band of Indian scouts, as recounted in numerous anecdotal histories. “We taught them a lesson they’ll ne’er soon forget,” he wrote.
Matheson later returned to his mine in the Sierra Madre, probably with a new mine crew to work on his treasured mine. Dobbs, Curtain and Traven may or may not have been among them (ok, I couldn’t resist. It is, after all, one of my all-time favorite books, and everybody loves the movie.) Matheson returned to Bisbee for supplies, only to find his best friend Tappenir murdered, a “brutal, cowardly murder.” Today, the event is known as “The Bisbee Massacre.” Matheson joined the second posse which captured two of the murderous gang. Five of the murderers were caught and hanged at the same time. The invitation to the hanging that is specifically addressed to Matheson is in the archive, as is his commentary. The sixth man was hung by the local Vigilance committee on a local telegraph pole a few days later.
So. That’s the short version. The letters had so much detail that much of it is far beyond what most or any anecdotal written histories seem to relate. In short, I found very little primary source material discussing the many events that unfolded and were an important part of this period of history. I even found it difficult to get a great description of Gen. Cooke’s expedition, though it may exist somewhere. Sometimes the informational “digging” takes far too long and can be costly. I thought I had to settle for what I had (still great stuff), until a new AI (Artificial Intelligence) APP (program on your computer or phone) was recommended. “You have to try it,” said my very experienced engineer friend. So I downloaded “Copilot.” We tried a number of queries that were science or engineering based, and it really did look pretty good. I was surprised. Then I tried the name I had spent so much time researching, “Robert Matheson.” Soon the query spat out a plethora of information, all of it claiming Matheson to be based in Arizona, and an author of a book. “Oh?,” I thought, “this is interesting.” How did it find out this stuff? Each line or segment was footnoted – all to the Arizona Historical Society files and to a book reportedly written by Matheson, “Mineral Resources of Arizona.” Little did I know at the time, that the AI program had probably mined most of the information from my own work on my own computer. But there were obvious mistakes. It called the famous Copper Queen Mine the “Copper Queen Gold Mining Co.” Matheson never worked for the Copper Queen. It gave the reference to the so-called Matheson book. But the book with a title that is most similar to that reported title was written by Patrick Hamilton (not Matheson), with several editions published (1883, 1884 in example.)
Resources of Arizona by Patrick Hamilton, 1884
It just kept getting worse… I dove into the footnotes, and dove into the online resource catalog of the AHS. I found … exactly … nothing. I had suspected the book was a fake because I’ve been deeply into Arizona mining history for five decades. That doesn’t mean there are things I haven’t discovered yet – there will be. But a reference book like that, that I haven’t read or heard about, and is not listed in any Arizona mining or geology bibliography? Phooey. So I wrote the AHS, thinking maybe I’m missing something. They got back to me immediately. Absolutely nothing on Robert Matheson, and they’d never heard of that book either. AI made the whole thing up.
The Aftermath
I brought it up the same day in a meeting that had several college and graduate students in attendance. “Oh, that’s too bad,” they said. “You might have to use more qualifiers or a different program to get what you want,” one person touted. “I use it all the time to write papers. I tell it to write in the voice of an xy year old student and use info from various sources.” … “And I get a pretty good paper.” I said, “If I were a professor, that paper would get an F. That’s why home school today doesn’t work. People cheat. I consider using AI as cheating. It needs to come from your brain from the learning process.” The response: “Its not cheating, its using new technology to get a better product.” My response; “It doesn’t come from your mind.” (the student walked away.)
When I mentioned this discourse with another colleague, he told me about all the problems with this exact situation, not only in grad students, but among faculty. …Oh my…
AI. Is it better for humanity, or is it the beginning of the destruction of the human intellect?
Holy smoke folks. AI has undoubtedly allowed for remarkable advances in science, engineering and technology. But now it is used to replace the very thought processes that make the true and future geniuses who come up with new concepts, theories, product designs, medical advances, etc.
I realize AI is part of the advancement of the human intellect. But to use it as a crutch, seems unimaginably wrong.
The Lesson
Be careful what you read. I’m not sure we will ever be able to be able to fully detect phony AI composed articles. I’m a fan of the old-fashioned primary source material. As I write this, I sit in a huge, private library of primary source material. I use it all the time. Each volume was composed by one or more authors and editors. Most is factual. Much is anecdotal, but of the time period. Modern data comes from the real, live, actual experimentation process, complete with reported successes and failures. Let’s hope humanity wakes up to the issues, and continues to use “the old noggin” to do the right stuff, not the fake stuff.
This essay was completely hand composed and is the sole product of my own thought processes based upon my 70-plus years of experiences on this earth.
Now… I wonder if I can use AI to help find the real Treasure of the Sierra Madre? Or the Lost Dutchman… or Quivera… or Cibola, the ancient “cities of gold”? …. Lemme see……
Vignette of miners in the Sierra Madre
Lot #4045 Matheson Letters, 1883-1886, "Bisbee Massacre" and Apache Attacks
Wild West Relics Auction April 2024