SETH W. JAMES

No AI Now, No AI Ever

July 11, 2026

By Seth W. James

It is a sad state of affairs that the below statement must be made, but as a matter of integrity and as a commitment to readers and to my fellow artists, I proudly pledge:

All of my previous books have been, and all of my future books will be, written solely by me, without the use of any so-called AIs, LLMs, or similar devices or programs.

Why is such a statement needed?  To put it as simply as possible, so-called AI is theft.

There is no such thing as Artificial Intelligence: it remains what it has always been, a matter of science fiction.  What is marketed as AI and, somewhat less dishonestly, as LLMs or Large Language Models, are nothing more than predictive models—a technology that has been around since the 1970s.  Most readers are already familiar with the outputs of such models, going back fifty years, in the form of weather reporting.  In this first, widely consumed model output, historical weather data—temperature, humidity, wind speed, etc—are gathered into a database, which is then queried by meteorologists to create a weather model, based upon those past data.  The model then focuses upon those data in-scope of the meteorologist’s query, assuming that the past will predict the future: if, say, 80% of the time, a thunderstorm over Boise correlates to a rainstorm over Chicago, three days later, when looking back over one-hundred years of data, the model will then return rain over Chicago as the prediction—because that is what exists in its database.

So-called AI works the same way when fabricating text: the huge, environment-, economy-, and health-destroying datacenters are the databases upon which the models draw, looking not only at weather data this time, but at as much data on as many subjects as can be packed in (at times, illegally or without notice to the original creators).  When a user enters a prompt—which is, in fact, a database query—the model returns a prediction result.  If you ask the so-called AI to write a review of a recent movie, the program will not watch the movie and offer an opinion—because it is not sapient and incapable of reason, it is simply a mathematical model—but, rather, the so-called AI will instead focus upon all movie reviews to which it has access, scraping their prose for the format of a movie review, lifting the content of other reviews of the movie queried, perhaps inserting copy taken from the movie’s tag line and description, and it then predicts what a review of that movie would look like.  The so-called AI is not writing a movie review: it is predicting what the writers within its datacenter would have written—using their words.

The negative effects are as obvious as they are nefarious.  In the first place, artists’ works are accessed—often without their consent or knowledge and without compensation or recognition—and reproduced, amalgamated with other stolen works to obfuscate their origins.  The accuracy of the so-called AIs’ predictions are now known to be highly suspect, if not flat-out nonsensical, leading to the term AI slop, but the negatives go deeper and last longer.  Because so-called AI accesses a finite database in its datacenters and produces a prediction based upon those data, new thought, new analysis, new assertions, art, direction, and experiences become impossible, only the endless reproduction of extant assets, the quality of which degrades with each prediction.  Like photocopying a picture and then repeatedly photocopying the resulting output, over and over again, quality and usefulness degrade to the point of total obliteration.

Perhaps more damaging to humanity than the cessation of new art, thought, and experience, is the hampering of humans’ abilities to reason, to think, and to create.  There are now legions of would-be authors and other creators using so-called AI to steal from real artists, using the prediction models to guess what their intended work would look like—and in the process steal the human-authored works of countless real people.  Beyond the degradation of those individuals and the rule of law, those would-be authors will now never grow, never develop the skills to actually tell the stories they would like to tell.  They are thieves, yes, but they are victims, too.

With a generation now going through their education while relying more and more on the flawed, continually degrading so-called AI prediction outputs, these students are not developing the reasoning abilities that would have otherwise empowered their voices, their works, and protected them against the falsehoods of domestic politicians and foreign adversaries—and the billionaires everywhere who hope to steal from us until there is nothing left.  I’m sure the billionaires look upon the dumbing-down of the world’s youth as a feature, not a flaw, of their predictive models.

In closing, I reiterate that all of my books are written by me and always will be.  I do not condone the use of so-called AIs, LLMs, or any other predictive models.  I encourage all aspiring artists to do the work necessary to developing their skills, their minds, and to truly become the artists they wish to be and to reject the easy path of theft.