Thomas Ptacek advocates for a strict rule when using LLMs for writing: never use a single word suggested by the AI. He proposes this as a form of intellectual personal protective equipment to maintain discipline and avoid the distinctive, often off-putting, style of LLM-generated text. While not using LLMs for content creation, Ptacek employs them for fact-checking, grammar correction, and as a thesaurus, sharing a prompt for a personal LLM copyediting tool. AI
IMPACT Suggests a disciplined approach to LLM integration in writing workflows, emphasizing human oversight and originality.
RANK_REASON Opinion piece from a named credible voice on how to use LLMs.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →