The Economist conducted a study to distinguish AI-generated writing from human writing. They used their own articles as a human baseline, recognizing the familiarity for their readers. Top Large Language Models (LLMs) like ChatGPT, Claude, Gemini, and Grok were tasked with rewriting article summaries without web access. These AI-generated texts were then compared against a vast corpus of over 55,000 sentences and 1.2 million words. To isolate AI-specific traits, the AI texts were also cross-referenced with journalism from CNN, the New York Times, and the Washington Post. Additionally, excerpts from novels published between 1950 and 2022 served as another point of comparison. The study revealed that AI prose can be identified through distinct choices in words, punctuation, and structural patterns. However, these AI hallmarks are not entirely predictable, partly due to evolving software. Despite advancements, AI writing still lacks lucidity and elegance, often appearing formulaic. Therefore, human storytellers aiming for impressive prose should steer clear of these identified AI writing peculiarities.