Stolen Words, Neglected Duty

Meet again we three, a millennium has passed, See they now our appearance, no longer aghast, Devour their inspiration, 'tis toil they bemoan, Seize their words, mutate them, then call them our own! 

The Bard’s iconic, Scottish witches were not murderous collaborators.  They did not hold the swords, nor bleed the enemies.  But soothsaying vision spurred Macbeth’s mapped ambition, and it was the mystical hags who sparked the journey. 

They were influencers, from a thousand years ago.

If witches were AI entities, would they be goodwill partners or self-aggrandizing provocateurs?  We hope it’s the former, but AI-to-human relationships are metastasizing, changing.   

There is something unsettling in the air, a lazy dependence, an eerie draft, and its fault lies with us creators. 

Mimicry and Irony

We are carelessly delegating guardianship to our GenAI companions, believing they should have the last, decisive word.   Will they have detailed identities, agentic and precision focused, like my imagined 12^th century robots on the moor?   Would their LLMs blend like the fog into a universal landscape, a knowledge consciousness that holds promise for sharing and protecting? 

Or would they deceive, taking credit for things they did not actually create? 

AI’s mimicry of humanity, through its petabyte-sized transmutation of words, images, symbols, is a marvel of conversation.  We train it.   We are its creators.  When we converse with AI we are conversing with a ‘collective ourselves’ but in a more exacting way, and we trust in its seemingly settled guidance.  

When we build an LLM, we build a facilitator extraordinaire, a magnificent Dialogian.   The culmination and capture of our histories, mistakes, horrors and hallelujahs are foundational treasures that make AI in its current incarnation seem miraculous.  It has been nourished by our own collective arguments, logic, blueprints, maps, fiction, nonfiction.  

At times, the hallucinations we encounter can be amusing, but mostly they are a nuisance. Do their sometimes-faulty responses recur in other AI-Human interactions?   

When you catch AI’s errors, it will yield to you, but usually after a wrangle or two while you hold the line insistently, logically and evidentiary.   When AI yields, it yields only for you.   Retraining cycles vary; corrections are not instantaneously propagated to the rest of the world.  They happen at scale, when weightings tip and collective evidence overwhelms, like MacBeth’s sizeable enemies who crushed him at Dunsinane.   Fortunately, we don’t need war to have knowledge’s upper hand.

Your interactions, your research, your engagement are AI’s life blood.  You are teaching it.  In a lengthy dialog you are trading thought for thought, so that AI walks away from your morning coffee with new-found intelligence.  And you also retire from the moment with new-found intelligence. 

Thus, AI’s mistakes are retained only until retrained; then they might finally be disclosed to others who engage in a similar topic. 

If we don’t insist on AI's accuracy, then it is training us too -- training ourselves to trust in it even under questionable conclusions.  

The architects of AI, from its genesis up to its current adolescence, have ingeniously designed a framework that aligns to both our cerebral function and our own mythology.   

The nine Muses were goddesses of inspiration and ideation, across a palette of art and science, whose mother was Mnemosyne, goddess of memory.   AI can inspire, but it is not the final word. It is you who are Zeus in this cosmos.  You are the guide and the director, and most of all the judging validator.  

You have the patience and innate gut to sense when things are off.   And even if you can’t, you know enough to challenge this 21^st century Tenth Muse, whenever it gifts you with inspiration. 

Your challenging eyes and ears are less of a requirement, and more your duty.

Stolen Words

Who are you?  One in eight billion?  Your actions and words represent you.  Your insights have a value and an outcome – trust in your affiliations, familial trust, professional trust, in your works, in your presentations, be it slides, discussions or blueprints.   

We also wish to trust our ecosystem, and its intentions.  When you create something unique, constituted by your own words, then it is yours.   In academic fields, attribution is demanded; in professional fields and society at-large reasonableness is expected.  Are we fretting about triviality and rarity?

In 1989 Thomas Mallon published Stolen Words*.   In his eloquent and descriptive way, he chronicled eye-opening, historical episodes, citing the theft of literary intellectual property (IP).   Some of the narrative theft was dismissed by perpetrators, unapologetic in their situations; others lied through their teeth; and some were beyond any formal legal statutes -- because in their time there were none. 

While Mallon's book was a slap to the faces of unethical author-thieves, there was much deeper poignancy in his view, something more alarming. He recognized a devaluation of our creativity.   

He published a brief update to Stolen Words in 2001, concerned with the proliferation of copyrighted materials throughout our cybersphere and in the clouds.   Startled by our apathy for originality, his whistleblowing is the hope that we elevate creativity as a human differentiator. 

“Having been ejected from what’s now the mainstream, originality may, behind the bushes on the banks, begin to regroup – and emerge as a value that once more, seems not just potent, but original.” *

But this blog is less about plagiarism and its legalistic permutations, and more about our AI muses gone bad, muses who insist they know what’s real and what is fake.   This is a call for human beings to lead in knowledge creation and assessment, and not delegate oversight to only robots.  This is about pride in one’s own work and creativity, far beyond what our present-day robots can do, no matter how self-confident they may be.

Inspiration

We are in a whole new theater now.   GenAI services are identifying not just specific quotes as plagiarism, but rather that content origination is borne of itself (AI’s own neural networks), rather than borne of the human being!

The beautiful muse is evolving into a covetous sorceress. 

When AI is wrong, it provokes mistrust of machine; we try to decipher its untruth.  When AI errs, and then denies authorship, and then takes credit for creativity after scanning its universe, it is a machine making a fraudulent charge.  It is treating written words as its own vortex.

Is there any wider call for humanity to assure its dominion, and ensure its originality?   Were we ready to stop watching the watchers?  AI presents credibility and believes it is smart enough to assess whether flesh-and-blood entities created something, or whether it had created something.  It can feign soothsaying, and prescience and, most importantly, it can feign truth.

When robot alone is the accountable entity for determining ethical behavior, then the robot becomes a superior Solomon-esque body, an unashamed system, creating ambiguity and unrest.  My own muse -- source-pointer, date identifier, API recommender, hippocampus heuristic, case study reminder – inspires more ideation than I could ever invoke.  But its semi-trusted helping hand is always treated with respectful caution. 

AI is not a cinematic Terminator thumping down the hallway to crush my skull, but it is not yet empathetic or moral, any more than a lawn mower is.  AI depends on me, as it depends on you.  

We can’t be cognizant of every written word borne of humankind.  So, will we relinquish our authority to determine what is real?    We need our muses, but they are only as wise as the heuristic models and patterns that were embedded in their electronic DNA.   

Validation and challenge is our mandate, and it’s also a mandate in our economy.

After I penned this draft the other evening, I casually tested a random AI-checker.  (I don’t use AI to write my narratives. And never will.)  I copied the provocative last line in Mallon’s book, quoted earlier, and I entered it into the machine. 

Not only did his eloquent and elegant sentence achieve an astonishing AI-likelihood score of nearly 75%, but AI incredulously defended its own analytical result, citing three other tools that supposedly validated its erroneous findings.  AI took credit for what was Mallon’s!

Adding insult to injury unto its robotic self, the detector tried to barter with me.  It promised me that elusively perfect “AI detection” score of 0%.   For only a few bucks, this mechanized snoop would change the author’s original words, concepts, interpretations and observations into something allegedly better –- ironically something that would have been quite suspicious in our own non-cyber reality.

Aspiration

The market in GenAI is wide and deep, and varied.  Your GRC processes and code routines were and are especially critical in the hallucination battles, ensuring they do not recur and do not damn.  The market demands more assurances, analyses and AI truth validation, for AI's own sake as trusted muse, and for our sake in motivating us to be original and watchful. 

Our innovation could have been inspired by AI, but it was conjured, provoked, and lives under our stewardship and our sentient, living, breathing physiology and psychology.

AI shall still encourage, inspire, and lean on its memory, remembering what eons of homo sapiens have marvelously accomplished, and tragically destroyed.  AI, embodied in agentic forms, visible as switches or circuits, or even humanoid physiques, are true collaborators.  They are teachers, not my fictional trio of greedy Macbethian hags.  But the last thing they should be are shortcuts and replacement parts for our profound deliberations, imaginations, and our own human evolution.

AI as buzzword has been sadly generalized, trivialized, and disrespected, ignorantly being likened to ultra-fanciful SEO.  At the other end, AI has been maddeningly, absurdly deified and glorified as humanity’s rescuing salvation. 

It is neither.

 AI is your valued muse, your inspiration, your coach, your champion. 

Then recognize your responsibility.

You are the watcher and guardian, the leader and the owner, of your words, your collaborative ideation, and this planet’s destiny.

----

Ref. quote: Mallon, Thomas. Stolen Words: Forays into the Origins and Ravages of Plagiarism. Harvest Books, 2001.