Claims Library Entry
The Em Dash Death Certificate
Kamil Banc explores how the em dash, a 300-year-old punctuation mark, became a telltale sign of AI-generated content due to overuse by language models. The article traces the mark's history, explains why AI overuses it, and offers practical alternatives like colons, semicolons, and varied sentence structure. The broader lesson is about recognizing and breaking predictable AI writing patterns to keep content authentically human.
Published August 31, 2025 by Kamil Banc
Lead claim
AI overuse turned the 300-year-old em dash into a predictable tell of machine-generated writing
Atomic Claims
What this article supports
Copy individual claims as needed.
Claim 1 · Source summary
Origins of the Em Dash
The em dash originated in the 1700s when printers measured it against the width of the letter M.
Claim 2 · Source summary
How AI Learned Dashes
Machine learning models learned to use em dashes after identifying them frequently in engaging conversational content.
Claim 3 · Source summary
Writers Abandon the Dash
Real writers started avoiding em dashes to dodge the AI label, killing a useful punctuation mark.
Claim 4 · Kamil's interpretation
Human Alternatives Recommended
Kamil Banc recommends colons, semicolons, parentheses, and full stops as human alternatives to overused em dashes.
Claim 5 · Kamil's interpretation
Edit AI Output Ruthlessly
Banc advises editing AI output ruthlessly, cutting every second dash, and varying sentence length.
Evidence
Context behind the claims
Quote
"The future belongs to writers who sound like humans talking to humans."
Key statistics
Alt + 0151
Windows keyboard shortcut for typing an em dash, which the author says he repeatedly had to look up
5-minute prompt technique
Banc's promoted method for transforming robotic AI drafts into conversational, human-sounding content
Every third sentence
The author's characterization of AI's predictable em dash placement pattern in generated text
Supporting context
The article blends historical narrative with practitioner observation rather than formal research, tracing the em dash from 1700s printing practices through its adoption by writers like Emily Dickinson to its current status as an AI writing tell. Banc's central argument is that machine learning models absorbed em dash patterns from engaging human content and now deploy them on a rigid schedule, making the punctuation a recognizable signature of generated text. For practitioners, the takeaway is practical: when using AI to draft content, ruthlessly edit output, cut repetitive dashes, vary sentence length, and replace formulaic transitions with line breaks. The broader principle extends beyond punctuation, since overusing any technique turns it into a detectable signature. Writers working manually should also self-assess whether their paragraph structures and transitions have become predictable habits.
How to Cite
Use the claim-level citation when you need a precise statement. Use the article or claims-collection citation when you want the wider argument and source context.
Individual Claim
Best when you need to cite one atomic claim directly inside a memo, deck, research note, or AI output.
"[claim text]" (Banc, Kamil, 2025, https://kbanc.com/claims-library/the-em-dash-death-certificate)Original Article
Use this when you want to cite the full newsletter article at AI Adopters Club rather than the structured claims page.
Banc, Kamil (2025, August 31, 2025). The Em Dash Death Certificate. AI Adopters Club. https://aiadopters.club/p/the-em-dash-death-certificateClaims Collection
Use this when you want to reference the full structured claims collection on this page.
Banc, Kamil (2025). The Em Dash Death Certificate [Structured Claims]. Retrieved from https://kbanc.com/claims-library/the-em-dash-death-certificateAttribution Requirements
- Include the author name: Kamil Banc.
- Include the source: AI Adopters Club or the structured claims page.
- Link to the original article or the claims page you used.
- Indicate any edits or transformations if you changed the wording.
Related Reading
More from the library
Where pure AI coding succeeds and where technical knowledge remains essential
5 claims
Experiential learning accelerates AI adoption
5 claims
Amazon's systematic AI implementation methodology
5 claims