Copying from the Scriptorium to the Cell Phone

The Science Fiction Writers of America publishes a news blog, “Writer Beware.” The July 31, 2026, entry is about changes to the Canon of Ethics of the American Association of Literary Agents. In their summary, the SFWA cites a comment from Publisher’s Weekly: 

“A surprising omission from the revisions is any language addressing AI. According to PW’s article on the revisions, though, this is something the AALA is looking at. Guidelines around AI usage are notably absent from this round of revisions, though AI use ranked as one of the top concerns among agents in a recent AALA survey. [AALA president Regina] Brooks said the AALA has begun circulating information internally and among members, but is “waiting for things to settle out” before formalizing anything.” 

Common chatter about AI is ubiquitous. I am complacent and ambivalent. I do recognize that soon (if not yesterday), an AI is going to sue in court for its rights as an artificial individual under the law, the same any other corporation. Whatever that may bring, for the immediate present, AI is just another computer program, a platform, an environment, no different than Cobol, Python, DuckDuckGo, or Amazon Web Services. Collectively, they are only wider abstractions to encompass concepts. 

It remains that the grammar checker and spellchecker are annoying but convenient. Typeahead is a klaxon for cliches. When you are in a Zoom call (or similar) turn on CC:Closed Captions and see whether and to what extent that program learns the nuances of your specialized vocabulary. 

“People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices.” – Adam Smith, The Wealth of Nations, (Book I, X “Wages and Profit,” Part I)

Writers and artists worry about losing income to new tools and it is an old problem. When movable type printing was invented copyists were disadvantaged; they rebelled and lost. When all copies were made by hand, it was a nice gift to copy a book and give it to your friend. The ability to produce hundreds or thousands of copies led to copyright laws. 

Until the invention of computers, copying and reading were distinct acts. In his 1985 book, Electronic Life, Michael Crichton asserted that reading with a computer is now identical with copying to a computer. My Apple iPhone connects automatically to my Apple iMac and copies content from one to the other.

During one of several continuing and ironic discussions with friends in the Fandom Association of Central Texas (FACT) about why we ban AI, the comment was made, “All they do is scrape the internet.” 

And what do you do? Does your story have a faster-than-light spaceship with colonists, or does it have dragons, vampires, witches, sorcerers, or does have robots or aliens? 

FACT sponsors the annual ArmadilloCon science fiction convention in Austin. During a Board meeting, discussing upcoming plans, it was said that an invitation to speak and present had been intended for a new popular writer but that as it happens, the writer-artist encourages “fanfic” written by their readers. Moreover, they actively promote and reward the use of AI tools to write and illustrate these offshoot narrations. So, the convention withdrew the invitation. To me, that was wrong of FACT.

On YouTube is a video about a monk learning to use a book. Danish is not required to understand the humor and it is interesting to read the cognates.

It is a cartoon that a proofreader found when my wife was an employee of a large commercial printer and also freelanced, reading over 100 books for Bantam Doubleday. The monk is running into the scriptorium with sweat and exclamation points over his head. He is barely carrying a double armload of scrolls. He says, "It should be celebrate!!"  

Science Fiction Writers of America https://sfwa.org/

SFWA “Writer Beware” blog by Victoria Strauss here: https://writerbeware.blog/2026/07/31/the-aala-makes-significant-updates-to-its-canon-of-ethics/

Publisher’s Weekly comment article here 

https://www.publishersweekly.com/pw/by-topic/industry-news/people/article/100919-major-aala-ethics-update-allows-agents-to-offer-paid-editorial-publicity-services.html

A Brief History of Printing.
Part II: The Economic History of Printing
by Frederick W. Hamilton,
Committee on Eduction, 1918.
From Gutenberg here.

"A Brief History of Printing. Part II: The Economic History of Printing" by Frederick W. Hamilton is a historical account likely written in the early 20th century. The work focuses on the economic dimensions surrounding the printing industry from the advent of printing in the mid-15th century through to the French Revolution in 1789. It delves into government regulations, labor conditions, and the evolving dynamics of the trade guilds impacted by the rise of printing. The opening of this volume introduces a thorough analysis of the printing industry’s economic landscape during a transformative period. It highlights how printers operated largely outside established guild systems, contending with governmental oversight and fluctuating market demands. Hamilton outlines how apprenticeships were structured, exploring the socioeconomic conditions faced by various classes of workers within the industry. He also contextualizes the historical significance of printing as not merely an industrial endeavor but as a pivotal influence on societal thought and subsequent revolutions, setting the stage for a detailed exploration of regulations, labor relations, and the inevitable shifts leading to modernization. (This is an automatically generated summary.)  A Brief History of Printing. Part II: The Economic History of Printing by Frederick W. Hamilton, Committee on Eduction, 1918.  

PREVIOUSLY ON NECESSARY FACTS

For the Glory of Old Lincoln High 

All Volitional Beings Deserve Rights 

Generative AI: As Big as it is, it is only a Program 

Art & Copy 

Start the Presses 

Hermann Zapf on Alphabets 

Libraries of the Founders 

Five Books about Books 

Goudy by Bernard Lewis and Goudy in His Own Words 

Frank Romano’s History of Desktop Publishing 

Generative Artificial Intelligence: As Big as it is, it is only a Program

We ignored the warnings from Mary Shelly, Thea von Harbou, Karel Čapek, and Colossus: The Forbin Project (1966) by Dennis Feltham Jones. The Hegelian dialectic predicts that the terror of the artificial human must have created its own contradiction and it did: the Apple Macintosh 1984 Super Bowl Commercial and everything it validated from Thomas Edison (“All I ask of my body is that it carry my brain around.”) to Alan Turing (he, him, G), Grace Hopper (0111), and the nerds of Silicon Valley. 

We made everyone a computer programmer. Some of us started earlier than others. We made it as easy as possible. 

For the people who never learned to program before, the user interface is as easy as a Google search or a Word document. Just ask. The program (measured against 1985) is an extremely large and complex compiler with a natural-language interpreter as the primary user interface. 

The Mercedes Maybach S 680 Brabus 
will not make coffee. It only does
extremely well what it was programmed to do.

The integrating truth which unites the two sides of the AI Dialectic to create the Synthesis which will result in a new Thesis, is that we all have been programming machines all of our lives. Aristotle never tuned a radio. Even given the wonderful opportunity, Plato would not get his hands dirty lubricating the powertrain of a steam engine. (Granted, there was Archimedes and it is a thin hypothesis that he built the Antikythera Device.) From learning to ride a bicycle, playing with Erector Sets, TinkerToys, and Legos and then learning to drive a car (and change a tire), we grew up interacting with machines.

I answered an ad from Murrary Resources, technical recruiters,
for a technical writer. "You do not look all that human to me,"
the human resources applicant tracking AI said to me.

A cogent comment on LinkedIn.
However, I replied:
"When the devil tempts you he will not come with fire and claws.
He will ask you to comprise a little bit just this once.  
Aaron Altman to Jane Craig in Broadcast News."
Because of Clippy, people stopped learning to write letters.

The Antikythera Device ran for a century or so, three lifetimes: created, admired, and never to be repeated. Scholars who study the device wonder and discuss whether and to what extent this was the work of one person. It has no simpler precursors, no antecedent subassemblies. The common informed wisdom includes the fact that this wonderful astronomical computer also, incidentally, was used to set the dates of religious ceremonies, which for the Greeks necessarily included stadium games. 

I ask: What if civic religion and social rituals were the purpose; and the path to that solution included the attendant abilities to track the planets? 

I suggest: It took a lifetime to build (three generations: one brainiac with minions), ran for a lifetime, and was lost in transport, falling into the sea for 2000 years. Would the loss of a single sword have benchmarked the existence of swords and how to make them? Many people could make swords. Not-many people built the Antikythera.

For 2000 years, the standard calculator was the abacus: no gears, no sundial, no pointers. It does have an accumulator and a register but no one ever called them that. 

GOSUB coffcoff. Despite their pretenses, programmers (now “devops”; formerly “data processing"), are just another class of users, like clericals and sales. They have other interfaces and presentations. Programmers will not tell you this, but they write with an “Application Development System” or “Program Developer Kit” which is a spellchecker for code. It is all colorized and color coded. If the programmer makes a mistake, the colors show that. Sometimes, the line itself will not accept an <Enter> if there is a bug on the line. That sure makes life easy. They now have “no code application development.” The programmers get to <quote> focus on higher level problems </quote> (ahem, coff-coff). RETURN.

Two hundred years ago, Charles Babbage built the Differential Engine and Lady Ada Lovelace programmed it. It took another lifetime for the world to catch on and catch up. The MIT “hackers” of the 1950s built model railroad switching systems from donated Bell Telephone racks. A hundred years ago, it had been suggested that in theory you could use your telephone to select and listen to a symphony orchestra in a distant city. 

Instead, we had broadcast radio—with commercials—an ethereal instantiation of newspapers, themselves an invention from only one previous century. 

Mathematicians complain about AI.

“First, it points out how AI models can “produce plausible but unreliable (or even incorrect) arguments which are difficult to distinguish from correct mathematical proofs.” Such developments put reviewers under increasing pressure and are “jeopardizing our ability to implement traditional standards for the correctness, transparency, and independent verifiability of proof,” the declaration warns.”

 — https://arstechnica.com/tech-policy/2026/06/mathematicians-warn-of-ai-threats-to-profession-as-industry-encroaches/

https://arstechnica.com/tech-policy/2026/06/mathematicians-warn-of-ai-threats-to-profession-as-industry-encroaches/#:~:text=“Mathematicians should find it quite,College London, in a statement.

However, no one spoke out against human mathematicians when Andrew Wiles announced his proof of Fermat's Last Theorem on 23 June 1993 at a lecture in Cambridge titled "Modular Forms, Elliptic Curves and Galois Representations". As it was told in Fermat’s Enigma: The Epic Quest to Solve the World’s Greatest Mathematical Problem by Simon Singh (New York: Walker, 1997; Anchor Doubleday, 1998), it was just another conference, but at the end of day one, the science journalists in the hall began calling their colleagues and by day three the place was packed. Andrew Wiles had proved Fermat’s Last Theorem. 

Or did he?

Again from the Singh work, closer inspection later by mathematicians at Wiles’s level revealed flaws in the proof. Wiles went back to work with a ream of white paper and a cup of sharp pencils. It took him three years. No one blamed him or his school or his parents. Instead, they gave him a medal. 

LET WORDS = 1024

The essential error in using AI (or a calculator) is seeking to avoid responsibility for choice and therefore not accepting responsibility for the consequences of your actions. Imagine that for an algebra class, you bought the instructor’s edition and copied out the answers to turn in for your homework. We recognize that as cheating. 

What if the book is wrong? Typos happen. The answer in the back can be wrong. A friend of mine had a math teacher who could not accept that the book was wrong. The kids did the problem correctly. Her book said that they were wrong. They met her after school (in the classroom, not the parking lot) and walked her through the answer. As it was told to me, the teacher refused to budge: the book could not be wrong. (She also called integers “intriguers”  perhaps indicating the challenge she found in high school maths.) So, AI can be wrong. Where is the surprise in that? 

Silicon Valley by Michael Rogers (Simon and Schuster, 1982).

"[Burt] Mathias is a self-made businessman who built the computer firm Solitron into a multimillion-dollar corporation. His college friend and partner, Alan Steinberg, is the computer genius of the operation, who explores the farthest reaches of computer intelligence while Mathias manages the finances. But Mathias stands on the brink of personal and professional ruin, about to lose both his wife and his business. To save Solitron from bankruptcy, he must gamble everything on whether Steinberg can perfect the first computer ever to simulate human consciousness. And once the job is complete, there is only one way to prove this remarkable ability to the incredulous world: the Turing Test..." (from the cover). 

As I remember the story, Mathias gamed the game by choosing as the human contestant, one of his own programmers, because who else would talk like a computer?

Big Bang Theory, NUMB3RS, NCIS,
Bones, Grey's Anatomy, replacing 
Leave it to Beaver, I Love Lucy, Lassie,
and Happy Days.

Clippy only does what it is programmed to do. The kids could prove their case, not just by working the problem but, had they chosen, via other paths. One of my physics professors, Dr. Alan Saaf at Lansing Community College, was answering homework questions at the blackboard. “I don’t understand number 3. …. How do you do number 5?... What equation do you use for number 1?...” He was going along and then he stopped. “You people would go out in the backyard and shoot hoops for 45 minutes and not make a single shot and still say you had a good time. How long did you spend on number 4? How many ways did you try to solve it?” 

There are over 300 proofs for the Pythagorean Theorem (one by Pres. James Garfield). The kids knew that the book was wrong because it contradicted known, provable truths. Clippy and Claude and the folk lack judgment.

This is an old problem among humans. In Computer Power and Human Reason: from Judgment to Calculation (1976), Joseph Weizenbaum warned of hackers whom he compared to compulsive gamblers. Driven by the superstition that one more patch will fix their problems, they stay up late, bleary-eyed and disheveled, working ever more frantically on a program they began without any reference to the substantive literature in the field in which they claim to be working. They are like gamblers who compulsively build complex rituals to control the game. 

Gamblers and Programmers

On the Cloudy Nights discussion board, which is mostly dedicated to chat about observational astronomy, in the forum for “Science! Astronomy, Space Exploration, and Others,” a topic title was the question “What can’t artificial intelligence do?”  The introductory post started: “We have made machines that can play chess better than we can. We are close to making machines that can write novels better than we can. Threshold question. Is there a limit? I can see no reason that there should be. The interesting question. What happens when we can make machines that can do everything better than we can?” In 100 replies, I was the only person who pointed out that while an AI could write a better novel, the novel itself was an invention. I received just one "like" for the comment. 

In a machine shop, there was a sign wrapped along the top of the walls: Good judgment comes from experience. Unfortunately, experience comes from poor judgment.  

For Cloudy Nights, I wrote: "(As far as we know) only humans can invent something new. You can say that an AI can write a novel better than a human, but the novel is an invention. As a form of narration and history, the novel is relatively recent. Poetry - epic poetry - was first. And before poems were invented, people made lists of things. ... Painting as we know it evolved in a series of quantum leaps. By the 4th century BCE graphical realism had achieved what we regard as modern techniques. The "Renaissance Masters" of Holland painted in a hyper-realistic style that violated "natural" vision. See The Arnolfini Portrait by Jan Van Eyck.  If you were in the room where the painter stood, you would not see the image in the mirror at the back the way it is presented in the painting. It is hyper-real. Impressionism, Expressionism, Abstract, ...  Performance Art.... That is the essential distinguishing characteristic that explains the difference between human intelligence and machine intelligence."

https://necessaryfacts.blogspot.com/2023/06/invisible-cheating-and-visible-rights.html

PREVIOUSLY ON NECESSARY FACTS

Documentation is Specification 

Denise Schmandt-Besserat: Accounting for Civilization

Denise Schmandt-Besserat: Art as Ordered Narrative

Love, Loss, and Redemption in Atlas Shrugged 

Atlas Shrugged Part 3: Who is John Galt

Remote Work Before Covid 

The Antikythera Device

Generative AI and Pseudographic Feynmans

The ability to train generative AI to impersonate humans is an example of something that can be done now because of new technology but which perhaps should never be done. 

I watched part of a video that is allegedly a lecture by Richard Feynman on why it will be impossible for explorers to Mars to return to Earth. In fact, there are many of these fake Feynman lectures on YouTube. They apparently do identify in the first seconds that they are an AI generated voice or graphic. They also match Feynman’s spoken delivery style, his turns of phrase and lecture pacing. I now appreciate the problem in textual forensics that confronted medieval scholars when they sorted out Pseudo-Aristotle and Pseudo-Dionysius.

By Greg Alleman from LinkedIn to YouTube.
Alleman has a career as an aerospace engineer.

These fake videos are intentional falsehoods which by their nature will replicate themselves particularly and others in general by creating an ecosystem that supports those lifeforms. Based on what I know about the evolution of life on Earth, I can only say that while some generative AI may be beneficial and others neutral, the impersonation of real humans is toxic. Videos from writers using generative AI will thrive creating and nourishing populations of pseudographia. It is inevitable that students who want to learn from The Feynman Lectures on Physics will land on one of these pseudo-Feynman lectures.

Two More Fake Feynmans

Yet Another Pair of Fake Feynmans

It is also true that each of us is responsible for determining the truth or falsehood of any assertion, even of our own assumptions. I never saw a P. F. Chang’s until I moved to Austin in 2011. I thought that it was local. I first saw a parody of Edward Hopper’s Nighthawks as a hip or campy wall poster back in the late 1960s before I learned the original in an art history class at Case-Western Reserve University in the summer of 1969. The first YouTube video I saw was Weird Al Yankovic’s “White and Nerdy” in 2005. A classmate in a criminal justice class at Washtenaw Community College showed it to me and I really identified with it. The kids in the class explained the other half of it. 

One Archive of Genuine Old Feynman Videos

I enjoy surfing YouTube videos because I like to admire the creative works of others and I learn a lot from them. Looking for printing and typography, I found Elle Cordova’s “Fonts Hanging Out” and “Punctuation Marks Hanging Out.” I also watch Mathologer and Numberphile and I have been going to sleep with David Miller’s series, Quantum Mechanics for Scientists and Engineers in my earphones. (I am testing hypnopedia. I will post here when I can speak the vocabulary.) I also subscribe to Rob Words, Audiomachine, Microbe Hunter, Jordan B. Peterson, and to Dwarkesh Patel interviewing Sarah Paine, among a dozen others.  

A friend who got tired of the AI-created misrepresentations of historical fact warned me about YouTube. Now I understand.

Granted the warning, these simulacrae could prove valuable. Perhaps we cannot have too many copies or reconstructions. Civilization is fragile and literacy is even more delicate. 

Aristotle’s own library was regarded as a royal treasure by the Macedonian ruling family. When they fell out amongst themselves, the scrolls were buried to keep them from competing claimants to the throne. When they were recovered, they were found to have been eaten through. Later scholars reconstructed them, but not well. Fortunately, those were not the only copies; Christian monasteries had preserved others. The Renaissance began when Petrarch and Dante visited those scriptoria to recover the lost writings of antiquity. However, they started with Latin, which they could read. Greek texts did not reblossom until the Renaissance took root in Venice, where Greek was known. Since then, scholars have improved the correctness of our reconstructions. In addition, other works have been unearthed. Aristotle’s lecture “On the Athenian Constitution” was rediscovered among the Oxyrhyncus Papyri. Similarly, in our time The Archimedes Palimpsest was recovered and lately one more page was found astray. But all of that only provides more evidence about what we know to be a huge body of lost works. “Oxyrhynchus yielded a huge random mass of everyday papers — private letters and shopping lists, tax returns and government circulars ... maybe 50,000 in all.” — https://oxyrhynchus.web.ox.ac.uk/waste-paper-city The Archimedes Palimpsest is another recovery with story of several losses and rescues**. 

(https://archimedespalimpsest.org/). So, these AI simulations could prove valuable.

The reason that I listen to Prof. David A. B. Miller’s lectures on quantum mechanics is that I found the textbook in the stacks at the Kuehn Math, Physics, and Astronomy Library when I was working at the University of Texas. It was perfect for me, a good midrange guide for college sophomores with clear narratives that are integrated with the relevant and necessary mathematics and instructive problems. I looked for a video and found the series. 

Image from Amazon. The book is no longer available
from Cambridge Press. Miller has a new volume out,
Modern Physics for Engineers and Scientists 
https://purl.stanford.edu/hr256mp8317

On that basis, imagine that some catastrophe has erased much of our collective electronic memory. Disastrous as the losses would be, they could not be complete. A later generation could scan the books and other printed media and also use AI to rebuild  the videos from fragments. The future could enjoy reconstructions of significant portions of our achievements while inevitably deploring the losses which they know only from inferences. 

** On 6 March 2026, the Zeitschrift für Papyrologie und Epigraphik reported the discovery of a lost page of the manuscript, now held by the Musée des Beaux-Arts de Blois. The text corresponds to page 123 of the Palimpset's version of the On The Sphere and Cylinder, Book 1, Proposition 39-41. The reverse depicts a painting of Daniel surrounded by two lions.— https://en.wikipedia.org/wiki/Archimedes_Palimpsest

PREVIOUSLY ON NECESSARY FACTS

About Richard P. Feynman

Genius by James Gleick 

Feynman's Rainbow by Leonard Mlodinow 

Observable Genius 

Merry Newtonmas 2019 

About Generative Artificial Intelligence

All Volitional Beings Deserve Rights

ArmadilloCon versus Artificial Intelligence

Invisible Cheating and Visible Rights

Digital Literacy and Artificial Intelligence