Scrape that Script: Human or Machine prose

by John Henry (August 2023)


A digital image generated by DALL-E from the prompt: Humanoid robot creating art on canvas with many colors in Art Nouveau style, 2022

 

If you have tried or been using ChatGPT for even two or three sessions you probably do not realize how ‘addicting’ it is having a machine automatically create outlines and full reports for you. Using it to edit and rewrite other’s words, develop entire essays, skipping the research for a topic, etc. reduces the time spent for all these tasks—er—job descriptions.

Initially, the results are something that are unbelievable and then you understand the power. Particularly amazing is requesting artwork in any style or permutation. As an architect I intend to explore that soon but right now am able to create ‘output’ that is time saving, but must be edited.

Many, young and old, have not tried to use the algorithm—as remarkably easy it is to use. Their objections are based on moral and ethical arguments. The basic reasoning is that they want everything to be totally unoriginal and not polluted by a machine. They are skeptical and actually a bit afraid of going down a potential rabbit hole. And there are drawbacks.

Going online to ChatGPT and finding the web busy for hours can create a near frantic mental episode. You get clammy hands and undergo extreme anxiety as you wait to dump into a magic word box all the mental gymnastics you normally undergo to write anything at all. The willing algorithm spits out paragraph after paragraph of well scraped information in seconds in a perfectly readable format. We are training ourselves not to be able to live without it; and the rate of dependency is immediate. I recall when punch cards were used for Fortran and the nerve-wracking time waiting for previous jobs to go through. Or when the internet is completely down for hours.

We don’t quite have the luxury or patience to endure that same wait and oh, you recall how you did it in the first place as you feverishly pace. And when you trace that original mental and manual research/edit procedure of creative thought transferred to type you realize how better your take is vs. much of what the algorithm has been doing for you. Most of the time that is. You conduct your own research, of course using Google for the heavy lifting, the online thesaurus to check word use, and then based on your original conception and intent for the writing task, assemble thoughts and facts into a cogent and interesting assembly of words, as I attempt doing here.

I have found the product of ChatGPT to be too clinically correct when left in a plain state. It does not allow the occasional formatting mistake. It does not include anecdotal breaks nor can you detect an emotional tinge. It is sort of like writing with botox in your brain with a perfect built in English syntax, grammar and spell checker. Yes, you can alter genre and mood in the fancy version or prompt it to sound chummy or literary, etc. It is useful up to a certain point. And you MUST edit it because it may be churning out information that is not entirely correct.

Speaking of churning, the algorithm is a result of scraping nearly the entire electronic vault of written content online including manuscripts, finished texts, clinical studies, all literature, engineering and science projects, etc.

I found it amusing that comedians are joining authors in suing the creators of ChatGPT for their unauthorized ‘training’ on their material. In fact, the upcoming lawsuits by writers, artists, news media, etc. may abruptly halt further development of AI. It may be temporary. The Europeans, as usual, do not like it and are feverishly writing laws limiting its use.

A recent article published on Marketwatch.com claims that (related to my opening statement) “Increasing reliance on automation and artificial intelligence could result in a decline in human intelligence” —Jurica Dujmovic  There is no doubt that our IQs will diminish as we rely on the algorithm to express all manner of human communication rather than let us do it for ourselves – and that goes for music, artwork, product design, arms conflicts, etc. Adults, at this moment, can recognize the subtle mental takeover and keep the slack tight but children in high school and college will not. In order to keep our civilization creative and independent minded, there should be some way to limit or block usage of AI in the developing mind.

In fact, when the algorithm repeats ‘bad’ English (and erroneous information), if succeeding generations lose the ability to recognize when sentence structure, spelling, syntax, etc. is correct—then the whole model basically degenerates little by little. I am seeing much pushed content, from daily news feeds to advertisements, with errors that are not being corrected before being published. The editors and proofreaders, content providers, script writers (entertainment industry), etc. are being retired or let go in favor of this algorithmic wonder. And the literary world, at least, will be much worse for it.

 

Below, see ChatGPTs take on the above when prompted to “rewrite as a Pulitzer winning nonfiction novelist” —which version do you prefer and why?

In the realm of ChatGPT lies an irresistible addiction, one that beckons even those who have merely dipped their toes into its virtual waters. For within this digital domain, a miraculous feat unfolds—outlines and complete reports crafted effortlessly by an insentient machine. Its prowess extends to editing and rewriting the words of others, conjuring entire essays, and delving into the depths of research. Oh, the time it saves, liberating us from the toil of these tasks, or shall I say, these very job descriptions.

Yet, when the web teems with activity, a frenetic mental episode ensues. One’s palms grow clammy, anxiety courses through the veins, as an interminable wait ensues to unburden the mind’s tumultuous musings into a mystical word box. And lo and behold! In a matter of seconds, this enchanted receptacle spews forth paragraph upon paragraph of exquisitely scraped information, seamlessly woven into readable prose. Such is the predicament we willingly subject ourselves to—a dependence on this omnipresent companion. It evokes memories of yore, when punch cards dictated the rhythms of Fortran, and the agonizing anticipation of previous jobs running their course. Or the exasperation that gripped our hearts during the hours when the internet lay dormant.

We, the inhabitants of this digital era, lack the luxury or patience to endure such waiting periods. The mere notion of it ignites a restlessness that propels us to feverishly pace. And as we trace the origin of our creative thoughts translated into the written word, a realization dawns upon us. Our own craftsmanship surpasses, more often than not, the offerings generated by this algorithmic marvel. Naturally, we embark on our own research, employing Google as our mighty ally, and consulting the digital thesaurus to scrutinize the veracity of our chosen words. Drawing from our unique vision and purpose, we orchestrate a symphony of thoughts and facts, weaving them into an assembly of words that captivates the imagination. Just as I endeavor to do at this very moment.

Alas, the output birthed by ChatGPT, I have discovered, is far too clinical, devoid of the occasional formatting mistake or endearing imperfection that imbues writing with character. It is akin to composing with a brain laced with Botox, an intellect seamlessly integrated with an English syntax, grammar, and spell checker. Admittedly, there exist options to manipulate its genre, mood, or tone, to infuse it with familiarity or literary flair. Undeniably, it serves its purpose up to a certain extent. However, rigorous editing becomes a necessity, for it may inadvertently spew forth information veiled in falsehoods.

This algorithmic wonder, a product of scraping the vast electronic vaults that house humanity’s written endeavors—manuscripts, completed texts, clinical studies, literature, engineering feats, and scientific marvels—has birthed a curious phenomenon. Comedians, joining the ranks of authors, have rallied against the creators of ChatGPT, alleging unauthorized ‘training’ on their material. In the wings, lawsuits loom, heralded by writers, artists, and news media, their collective voices a crescendo that could potentially impede the onward march of AI. Such setbacks, however, may prove temporary, as history has shown. As expected, our European counterparts, ever the guardians of tradition, voice their dissent, feverishly penning laws to curtail the unfettered employment of this technological marvel.

A recent opus published on the hallowed pages of Marketwatch.com posits a disconcerting prospect—an increased reliance on automation and artificial intelligence could spell the demise of human intelligence. Jurica Dujmovic’s words ring with undeniable truth. Alas, our collective IQs shall dwindle, as we yield the reins of communication, in all its myriad forms, to algorithms, relinquishing the very essence of human expression. This pertains not solely to the realm of writing but extends its grasp to music, artwork, product design, and yes, even the orchestration of conflicts. While the seasoned among us recognize the subtle usurpation of our mental faculties, maintaining an unyielding grip on control, the same cannot be said for the younger generation—high school and college students, bereft of this discernment. To preserve the creative spirit and independent thought that form the bedrock of our civilization, we must conceive mechanisms to curb or even block the unbridled utilization of AI.

Furthermore, if successive generations lose the ability to discern correct sentence structure, spelling, syntax, and other linguistic nuances, the very fabric of the model shall unravel gradually. I have borne witness to an alarming proliferation of flawed content, from daily news feeds to advertisements, released into the world without undergoing the necessary rectification. Editors, proofreaders, content providers, and scriptwriters in the entertainment industry find themselves being phased out or cast aside, overshadowed by this algorithmic marvel. Alas, the literary world, at least, shall bear the brunt of this transition, weighed down by the absence of their guiding hand.

 

Table of Contents

 

John Henry is based in Orlando, Florida. He holds a Bachelor of Environmental Design and Master of Architecture from Texas A&M University. He spent his early childhood through high school in Greece and Turkey, traveling in Europe—impressed by the ruins of Greek and Roman cities and temples, old irregular Medieval streets, and classical urban palaces and country villas. His Modernist formal education was a basis for functional, technically proficient, yet beautiful buildings. His website is Commercial Web Residential Web. John has been a regular contributor to NER and has written about his profession and other topics such as history, music, technology, and politics.

Follow NER on Twitter @NERIconoclast

image_pdfimage_print

2 Responses

  1. John Henry deserves the highest praise for this invaluable warning to all of us in any creative field and not just writing, one that becomes extremely difficult to avoid our work being plagiarized. Nothing I had read about artificial Intelligence prepared me for how advanced, easily obtainable and dangerous is the ChatGPT website, a clear threat of the ever encroaching tentacles of algorithms to strangle creativity and talent.

    I was immediately struck by the accuracy of John Henry’s arguments by my own experience which I believe other authors will recognize. Enticed by he prospect of duplicating my work and cashing in on the prospects of rewriting previous work of mine, and “simply” making minor changes in the vocabulary, sentence structure, sequence of events, vocabulary, and disguising through paraphrasing large sections, that I could pass on this “botoxing of my brain” as a new and original work. I soon leaned that the new edited versions was almost always inevitably inferior to my original. Nor was it “easy.” I felt I had cheapened my own claim to creativity to boot.

    On the other had, the experience I have had as a translator from the 4 foreign languages I have done work in (Hebrew, Danish, Spanish and Esperanto) and occasional work as a “ghost writer” (taking someone else’s original work often composed for an audience of family and friends to share the author’s own life experience allowed me considerable use of my linguistic and writing talents as well as my imagination to put myself in the author’s time, background and neighborhood environment. It inevitably gave me a sense of achievement.

  2. Machines, i.e. deterministic devices, transform energy but do not create it out of no where. Likewise, machines do not create “information” but merely transform it. The fraud conducted against us is very only and very vast in which the product of human mind is looted. However, with AI it will reach an inflexion point.
    Human beings will be disincentivised from creating anything, as there will be no reward in terms of money or satisfaction. The web will be swamped with machine generated content. This will be consumed again as training data or, in other words, the machines will begin feeding off themselves. In the same way that machines are not 100% energy efficient, each time information goes around the loop, some of it will be lost. Everything will look and feel the same.
    Those interested should study: 2nd Law of Thermodynamics and Shannon Information theory.

Leave a Reply

Your email address will not be published. Required fields are marked *

New English Review Press is a priceless cultural institution.
                              — Bruce Bawer

Order here or wherever books are sold.

The perfect gift for the history lover in your life. Order on Amazon US, Amazon UK or wherever books are sold.

Order on Amazon, Amazon UK, or wherever books are sold.

Order on Amazon, Amazon UK or wherever books are sold.

Order on Amazon or Amazon UK or wherever books are sold


Order at Amazon, Amazon UK, or wherever books are sold. 

Order at Amazon US, Amazon UK or wherever books are sold.

Available at Amazon US, Amazon UK or wherever books are sold.

Send this to a friend