Five Rules for AI Writing
The AI Daily Briefgo watch the original →
the gist
AI writing is a permanent fixture in professional workflows, but its utility depends on matching the tool to the task and avoiding the trap of outsourcing critical thinking.
The Shift in AI Writing Norms
The recent controversy surrounding Stanley Druckenmiller’s AI-generated Wall Street Journal op-ed highlights a shift in professional standards. While critics argued that using AI for high-stakes opinion pieces constitutes plagiarism or a lack of sincerity, the broader consensus among industry figures is that AI is a functional tool for articulation. The core issue is not the use of AI itself, but the perception of laziness when authors publish unedited, generic AI output—often characterized by repetitive syntax and "AISMs"—which suggests a lack of intellectual rigor in the underlying argument.
Five Rules for AI Writing
- Context-Specific Rules: Different formats require different approaches; an email, a strategy memo, and an op-ed serve distinct purposes and demand varying levels of human oversight.
- Quality Over Purity: The "purity test" of whether AI was used will fade, but the "quality test" remains; increased ease of input places a higher burden on the final output to be compelling and original.
- Effort Perception: Readers equate effort with quality. Glaring AI-generated syntax errors signal that the author did not care enough to refine the work, which undermines the credibility of the argument itself.
- Brevity is Superior: AI models are often trained to be verbose, but effective writing is concise. Following the principle that "I have not had time to make it shorter," authors must aggressively edit AI drafts to remove fluff.
- Writing is Thinking: The process of defining a thesis and constructing a narrative is an intellectual exercise. Using AI to generate the final text is acceptable only if the author has already performed the heavy lifting of thinking through the core strategy and logic.
Practical Application
For tasks like meeting summaries, AI is effective for compression, though it often misses the most important insights, requiring a human to add context. For strategy memos, AI should be used only after the strategy is fully formed, using the tool for outlining and iteration rather than initial drafting. Social media copy benefits from AI for short-form content, but requires significant voice-tuning to avoid sounding generic, and it cannot replace the human engagement required to build an audience.