Observed Signal · Sep 3, 2026 · Analysis · Source: Storytelling Edge · Impact: 2/5 · Sentiment: Neutral
Why People Mistake Em-Dashes for AI Writing
Storytelling Edge's newsletter explores why readers increasingly associate em-dashes and polished prose with AI-generated writing. It references a Wired story on a growing 'anti-AI literary counterculture' in which writers avoid AI conventions like negative parallelism and deliberately introduce small mistakes to signal human authorship. The piece cites an Economist study showing ChatGPT now rarely uses em-dashes, while Claude still uses them more than human writers. Linguist Gretchen McCulloch explains the linguistic concept of enregisterment, linking the em-dash's stigma to learned associations rather than measurable overuse by AI. The newsletter also discusses Substack's AI-detection feature built with Pangram, citing University of Chicago and UPenn studies that claim over 99% accuracy, and notes Pangram CEO Max Sperro's unclaimed $100 bounty for false positives. The author concludes that writers should develop a distinct personal voice rather than simply avoiding AI tells.
Explores the cultural and linguistic impact of AI-generated text on writing, including Substack's AI-detection partnership with Pangram and new research on AI writing markers; relevant to content and publishing but not a major AdTech industry event.
Track Substack Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Wired reported on a growing 'anti-AI literary counterculture' in which writers avoid AI writing conventions and deliberately leave mistakes to prove human authorship.
- The Economist published a study finding ChatGPT barely uses em-dashes, while Claude uses them more frequently than human writers.
- Substack introduced an AI-detection feature in partnership with Pangram to scan posts for AI-generated text, aiming to fight 'Claude-fishing'.
- Citing University of Chicago and UPenn studies, the article says Pangram is over 99% accurate with almost no false positives.
- Pangram CEO Max Sperro offered a $100 bounty for evidence of false positives; the article says nobody has claimed it.
Connected Companies & Entities
5 Entities mapped“In July, Substack introduced a new feature in partnership with AI-detection tool Pangram to scan posts for AI-generated text to fight 'Claud...”
“Substack introduced a new feature in partnership with AI-detection tool Pangram to scan posts for AI-generated text.”
“Wired recently published a fascinating story about how writers are creating a new 'anti-AI literary counterculture'.”
“OpenAI CEO Sam Altman has long been on a mission to get AI to stop using so many em-dashes.”
“According to a new study by The Economist, only Claude now uses em dashes more than human writers do.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
AI Writing Vibe Shift Accelerates
The newsletter analyzes recent signals showing a shift in attitudes toward AI-generated writing: Substack is partnering with Pangram to detect AI-written posts, LinkedIn introduced a 'Seems Like AI Slop' feedback button, Anthropic announced an invisible watermark (and a watermark API) for Claude outputs to meet EU transparency rules, publishers and agents are drawing lines on suspected AI-authored manuscripts, and several tech companies and marketing teams are instituting internal guidelines discouraging AI-written copy. The piece frames these moves as part of a broader industry move toward transparency and protecting brand voice.
Why AI Struggles to Write with Human Voice
The newsletter argues that large language models (LLMs) produce technically competent but emotionallyflat and generic writing because of how they are trained and fine‑tuned. Citing Jasmine Sun’s Atlantic piece and a Google DeepMind paper, the author says LLMs are trained on vast, noisy datasets and then tuned to prioritize safe, commercially valuable outputs—favoring corporate communication over distinctive, 'weird' voices. As a result, AI writing often lacks lived experience, evocative metaphor, and stakes. The piece recommends using AI for neutral tasks (press releases, landing pages, memos) but not for imaginative storytelling, and points to evidence that AI assistance can neutralize and erode an author’s original voice.
AI Writing Is Drowning the Internet; Pangram Fights Back
In an interview, Derek Thompson and Pangram CEO Max Spero discuss the flood of AI-generated content ('slop') on the internet, the rise of detection tools, and the implications for trust, media, and authenticity. Key findings include that 40% of webpages may be AI-influenced, and Pangram's detector boasts near-zero false positives. The conversation covers Pangram's methods, its integration with Substack, and the risk of big tech entering the space.
Receive all future signals for Substack & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.