Repeat something often enough, and people will begin to believe it. Like gossips hanging over the garden fence and sharing a juicy bit of information about someone or something, until they themselves begin to believe the outlandish ‘FACTS’.
Regurgitating Information
AI is no different. With each retelling or sharing in the chat rooms, a story takes on a life of its own and AI—not known for its reasoning, but only reacting—absorbs these chatter as fact-based. A case in point is where a recent post shared on Reddit was cited by AI when called upon. A parody post claiming X was cited as fact by AI, despite being clearly labelled satire.
We even find it in poetry, where forms are misrepresented across platforms. The Diminished Hexaverse is often defined as a five-stanza poem, but the prefix hexa- clearly signals six. What’s being described is a pentaverse, not a hexaverse. This misstep shows how easily repetition erases original intent. The inventor’s design was six words in the first line, then five, four, three, two, and finally one (supplying the poem’s title)—linguistic play on hexa (six). Not five syllables or stanzas, but a form of reverse rhopalic verse, where the line length diminishes word by word.
Another example is where AI habitually produces Suzette Prime with 9-syllable lines—despite the fact that 9 is not a prime. This happens because the internet is saturated with incorrect examples of Suzette Prime poetry. AI finds the common pattern and faithfully regurgitates what it scans, not what is correct. Since AI weights repetition heavily, if most examples online are wrong, the wrong pattern becomes dominant. Yet the form’s scaffolding is philosophical with prime numbers as building blocks. Nine, being composite, breaks the intent. It is not experimental poetry, but failed Suzette Prime poetry. This should be one of the red flags when judging a Suzette Prime contest—no magic trick, just common sense.
A Refection in Suzette Prime
I search here
I search there
I search everywhere
no poem in sight
they burn books
ban the rest
as fear guide their hand
if a stone should float
it’s a witch
but if a fish floats
we’ll be unwise to eat it
© Suzette Richards (2026)
Today, we can instantly recognise the absurdity of the ‘floating stone’ test used to condemn witches of old. The rationale behind that was if the subject floated in a river, they were guilty as charged; if they sank like a stone, they were innocent—scant consolation for the drowned victims. This shows how fear can override critical thinking, turning nonsense into deadly judgment.
AI Hallucinations
It is harder to spot AI hallucinations, because they take different forms: fabricated citations, misattributed facts, or confidently delivered errors that dissolve under scrutiny.
When AI misconstrues satire as fact, or reproduces the misdefined ‘five-line hexaverse’, it is hallucinating—confidently generating a false version of the form based on flawed online repetition rather than the original six-word descent. Similarly, when producing Suzette Prime with nine syllables in a line, it is hallucination as AI invents a ‘rule’ that doesn’t exist, because it misreads or mangles scarce references.
AI repeats legends and simplifications, and in doing so, it erodes authenticity.
Bias vs Hallucination in Poetry Forms
Bias: This happens when AI leans toward the most frequently repeated version of a form in its training data. For example, the 5-7-5 haiku ‘rule’ is so dominant online that the system defaults to it. That’s bias: a statistical tilt toward what’s common, even if it’s oversimplified or misleading.
Hallucination: This occurs when AI generates something that is not grounded in reality at all—confidently inventing or misrepresenting. For instance, calling the Hexaverse a five-line form despite the prefix hexa- meaning six. That’s not just bias toward repetition; it’s a fabricated mismatch between linguistic logic and output.
A Remedy
Your best defence is to hone your skills and fine-tune your nonsense metre. Gossip and misinformation have always been part of our lives—some relish the chaos it brings, without reckoning. It is easy to say: ‘But it has always been done this way’.
Humans and AI share the same flaw: the tendency to confuse echo with truth. The difference is that humans choose whether to check sources, while AI simply mirrors what it has absorbed.
Contest sponsors are in the same position as AI systems: they need to cross-reference the source of the poetic form they’re featuring, rather than blindly repeating what’s most common online.
Having a working knowledge of a subject gives you the foundation, but it’s the inquisitive instinct—the willingness to investigate the nuances of fact versus fiction—that keeps one’s work sharp and credible.
Conclusion
Poets who value authenticity rather than pandering to popular shortcuts will ultimately have the satisfaction of watching the AI-dependent crowd fade into irrelevance.
AI is never the source—just the echo chamber, like the village gossip. As YOU are ultimately responsible for what you post, it is wise to think about the source and not to blindly follow where no reasoning had gone before.
Image was AI-generated (1/7/2026)