Nathaniel Whittemore uses the controversy over an AI-assisted Wall Street Journal opinion piece to separate the legitimacy of AI writingAI writing is the use of AI tools to research, organize, draft, revise or evaluate written content while a person remains responsible for the purpose, accuracy and final voice. from its quality. He argues that AI assistance is becoming normal, but readers will still judge whether the author supplied genuine ideas, credible judgment and enough effort to make the work worth reading.
Nathaniel Whittemore proposes five rules: apply different standards to different kinds of writing, expect purity tests around AI use to fade, recognize that readers use perceived effort as a proxy for quality, prefer concise output, and keep ownership of the thinking beneath the prose. He frames AI as strongest in the middle of the processHuman-AI collaboration combines human judgment and accountability with AI speed, generation, analysis, and tool use in a shared workflow., where it can research, organize, draft and challenge ideas after a person has defined the thesis and before that person verifies the result.
The appropriate division of labor changes by format. AI can safely compress meeting notes and help with routine email, while strategy memos and persuasive essays demand much more human reasoning and review. For social posts and marketing copy, Nathaniel Whittemore recommends extensive voice examples, realistic expectations and using models to generate options rather than treating their first draft as finished workAI model output is the text, image, audio, code, prediction or other result a model produces from its input and learned patterns..
The practical conclusion is an effort sandwich: think clearly before promptingPrompt engineering is the practice of designing and refining instructions, context, examples, and constraints to obtain useful AI outputs., use AI to accelerate the middle, then edit for accuracy, brevity, voice and obvious model mannerisms. AI can make writing faster and more accessibleAI-generated content is text, imagery, audio, video, code, or other material produced substantially by a generative AI system., but it cannot remove the need to form a specific argument or care about the audience.
Watch on YouTube
