AI Writing Tropes to Avoid
AI Writing Tropes to Avoid
Word Choice
"Quietly" and Other Magic Adverbs
Overuse of adverbs like "quietly," "deeply," "fundamentally," "remarkably," and "arguably" to convey subtle significance. AI employs these to elevate mundane descriptions.
Examples to avoid: "quietly orchestrating workflows, decisions, and interactions" or "a quiet intelligence behind it."
"Delve" and Friends
"Delve" became an infamous AI indicator. This category includes overused vocabulary: "certainly," "utilize," "leverage" (as verb), "robust," "streamline," and "harness."
Problematic patterns: "Let's delve into the details..." or "We certainly need to leverage these robust frameworks..."
"Tapestry" and "Landscape"
Ornate nouns replace simpler alternatives. "Tapestry" describes anything interconnected; "landscape" labels any field. Related offenders: "paradigm," "synergy," "ecosystem," "framework."
Patterns to avoid: "The rich tapestry of human experience" or "Navigating the complex landscape of modern AI."
The "Serves As" Dodge
Replacing basic "is/are" with pompous alternatives: "serves as," "stands as," "marks," or "represents."
Examples: "The building serves as a reminder of the city's heritage" or "Gallery 825 serves as LAAA's exhibition space."
Sentence Structure
Negative Parallelism
The "It's not X -- it's Y" pattern, often using em dashes. Frames everything as surprising reframes, creating false profundity. Includes variants like "not because X, but because Y" and cross-sentence reframes.
Avoid: "It's not bold. It's backwards" or "Half the bugs you chase aren't in your code. They're in your head."
"Not X. Not Y. Just Z."
The dramatic countdown pattern building tension before revealing the actual point.
Examples: "Not a bug. Not a feature. A fundamental design flaw" or "Not ten. Not fifty. Five hundred and twenty-three lint violations across 67 files."
"The X? A Y."
Self-posed rhetorical questions answered immediately in the next sentence for dramatic effect.
Patterns: "The result? Devastating" or "The worst part? Nobody saw it coming."
Anaphora Abuse
Repeating the same sentence opening multiple times in succession.
Avoid: "They assume that users will pay... They assume that developers will build..." or "They could expose... They could offer... They could provide..."
Tricolon Abuse
Overuse of the rule-of-three pattern, often extended to four or five items.
Examples: "Products impress people; platforms empower them. Products solve problems; platforms create worlds" or listing "identity, payments, compute, distribution."
"It's Worth Noting"
Filler transitions signaling nothing, used without actual connection to previous arguments. Also includes: "It bears mentioning," "Importantly," "Interestingly," "Notably."
Problematic use: "It's worth noting that this approach has limitations" or "Importantly, we must consider the broader implications."
Superficial Analyses
Tacking present participle phrases ("-ing") to inject shallow analysis. Examples: "highlighting its importance," "reflecting broader trends," or "contributing to the development of..."
Avoid: "contributing to the region's rich cultural heritage" or "underscoring its role as a dynamic hub."
False Ranges
Using "from X to Y" constructions where items aren't on any real scale or spectrum.
Examples: "From innovation to implementation to cultural transformation" or "From the singularity of the Big Bang to the grand cosmic web."
Paragraph Structure
Short Punchy Fragments
Excessive standalone very short sentences or fragments for manufactured emphasis. RLHF training pushes models toward one thought per sentence with no mental effort required.
Avoid: "He published this. Openly. In a book. As a priest" or "These weren't just products. And the software side matched."
Listicle in a Trench Coat
Numbered or labeled points disguised as continuous prose using "The first... The second... The third..." openings.
Examples: "The first wall is the absence of a free, scoped API... The second wall is the lack of delegated access..." or "The second takeaway is... The third takeaway is..."
Tone
"Here's the Kicker"
False suspense transitions promising revelations that deliver unremarkable observations. Also: "Here's the thing," "Here's where it gets interesting," "Here's what most people miss," "Here's the deal."
Avoid: "Here's the kicker" or "Here's the thing about AI adoption."
"Think of It As..."
The patronizing analogy defaulting to teacher mode, assuming readers need metaphors for understanding.
Examples: "Think of it like a highway system for data" or "It's like a Swiss Army knife for your workflow."
"Imagine a World Where..."
Classic AI futurism invitation beginning with "Imagine" followed by wonderful outcomes.
Avoid: "Imagine a world where every tool you use has a quiet intelligence behind it."
False Vulnerability
Simulated self-awareness or honesty that reads as performative. Real vulnerability is specific; AI vulnerability is polished and safe.
Examples: "And yes, I'm openly in love with the platform model" or "This is not a rant; it's a diagnosis."
"The Truth Is Simple"
Asserting obviousness instead of proving it. Includes dramatic reveal variant: "The real story is..." while waving away everything prior.
Avoid: "The reality is simpler and less flattering" or "History is unambiguous on this point."
Grandiose Stakes Inflation
Everything becomes world-historically significant. A blog post about API pricing becomes meditation on civilization's fate.
Examples: "This will fundamentally reshape how we think about everything" or "will define the next era of computing."
"Let's Break This Down"
Pedagogical voice assuming hand-holding needed, even for expert audiences. Also: "Let's unpack this," "Let's explore," "Let's dive in."
Avoid: "Let's break this down step by step" or "Let's unpack what this really means."
Vague Attributions
Attributing claims to unnamed authorities: "experts," "observers," "industry reports," without specificity or actual sourcing.
Examples: "Experts argue that this approach has significant drawbacks" or "Industry reports suggest that adoption is accelerating."
Invented Concept Labels
Clusters of invented compound labels sounding analytical without grounding. Appends abstract nouns like "paradox," "trap," "creep," "divide," "vacuum," "inversion" to domain words.
Avoid: "the supervision paradox," "the acceleration trap," or "workload creep."
Formatting
Em-Dash Addiction
Compulsive overuse of em dashes for dramatic pauses, asides, and pivots. Humans use 2-3 naturally; AI uses 20+.
Avoid: "The problem -- and this is the part nobody talks about -- is systemic" or "The tinkerer spirit didn't die of natural causes -- it was bought out."
Bold-First Bullets
Every bullet point starts with bolded phrases. Extremely common in Claude and ChatGPT markdown. Almost nobody writes this way naturally.
Examples: "Security: Environment-based configuration with..." or "Performance: Lazy loading of expensive resources..."
Unicode Decoration
Use of unicode arrows (->), smart/curly quotes, and special characters not easily typed on standard keyboards. Real writers produce straight quotes and standard arrow characters.
Avoid: "Input → Processing → Output" or using curly quotes instead of straight quotes.
Composition
Fractal Summaries
"What I'm going to tell you; what I'm telling you; what I just told you" applied at every document level. Every subsection, section, and document gets summarized repeatedly.
Avoid: "In this section, we'll explore... [later] ...as we've seen in this section" or "And so we return to where we began."
The Dead Metaphor
Latching onto a single metaphor and repeating it 5-10 times throughout. Humans introduce, use, then move on.
Examples: "The ecosystem needs ecosystems to build ecosystem value" or repeatedly using "walls and doors" throughout an article.
Historical Analogy Stacking
Rapid-fire listing of historical companies or tech revolutions to build false authority, particularly common in technical writing.
Avoid: "Apple didn't build Uber. Facebook didn't build Spotify. Stripe didn't build Shopify" or "Take Spotify... Or consider Uber... Airbnb followed a similar path..."
One-Point Dilution
Restating a single argument in 10 different ways across thousands of words, padding simple theses with different metaphors and examples.
Examples: "The same point, restated eight ways across 4000 words" where each section rephrases without adding value.
Content Duplication
Repeating entire sections or paragraphs verbatim within the same piece, indicating lost tracking of what's already written.
Avoid: Identical sections appearing twice or paragraphs with identical wording.
The Signposted Conclusion
Explicitly announcing conclusions with "In conclusion," "To sum up," or "In summary." Competent writing doesn't need these signals; readers feel the conclusion naturally.
Avoid: "In conclusion, the future of AI depends on..." or "To sum up, we've explored three key themes..."
"Despite Its Challenges..."
Rigid formula acknowledging problems only to immediately dismiss them, following predictable beats.
Examples: "Despite these challenges, the initiative continues to thrive" or "Despite its industrial and residential prosperity, [location] faces typical urban challenges."
Key Principle
Any single trope used once might be acceptable. The problem emerges when multiple tropes appear together or when individual tropes repeat. Authentic writing demonstrates variation, imperfection, and specificity rather than algorithmic patterns.