All Topics

Technology

Art

The Curious Case of AI Language Models' Obsession with Em-Dashes

ahamez

7mo ago· 11 min readenInsight

100/100

Golden Brown

Bagelometer↗

The kind of bagel that ruins lesser bagels for you.

Score100TypeanalysisSentimentneutral

Summary

The article explores the curious phenomenon of AI language models' excessive use of em-dashes in their writing. It examines why this punctuation mark has become a defining feature of AI-generated text, noting that real humans who previously used em-dashes have stopped doing so to avoid being mistaken for AI. The piece discusses various theories about why models favor em-dashes, including training data patterns and the punctuation's functional versatility, while also noting the difficulty of prompting models to avoid using them. The article presents this as an interesting linguistic quirk of AI systems that lacks a definitive explanation.

Key quotes

· 4 pulled

If you asked most people to name a defining feature of AI-generated writing, they'd probably say the em-dash — like this.

Language models use em-dashes so much that real humans who like em-dashes have stopped using them out of fear of being confused with AI.

It's also surprisingly hard to prompt models to avoid em-dashes: take this thread from the OpenAI forums where users share their unsuccessful attempts.

Given all that, it's kind of weird that we don't really know why language models use the em-dash so much.

Snippet from the RSS feed

If you asked most people to name a defining feature of AI-generated writing, they’d probably say the em-dash — like this. Language models use em-dashes so much that real humans who like em-dashes have stopped using them out of fear of being confused with

You might also wanna read

Undash Chrome Extension Removes Em Dashes from AI-Generated Text

Undash is a Chrome extension that automatically removes em dashes (—) from AI-generated text when pasting into websites, social media, email