How a Professional Writer Uses AI to Compound Productivity
the gist
Writer Katie Parrott explains how she uses AI not just for drafting, but as a 'content agency of one' by building context-heavy systems that allow her to focus on high-level insights rather than repetitive execution.
The Evolution of AI as a 'Tool for Thought'
Katie Parrott’s journey with AI began as a necessity during a period of unemployment and mental health struggles. Unable to afford a professional career coach, she used ChatGPT as an externalized thinking partner. This experience shifted her perspective from viewing AI as a simple text generator to seeing it as a 'tool for thought' that could provide structure, accountability, and clarity. She emphasizes that AI does not provide answers gift-wrapped; rather, the user must engage in the hard work of questioning and prompting to extract their own latent insights.
Building a 'Content Agency of One'
Parrott describes her transition into a high-output writer by treating AI like a content agency. The core of her system is 'context engineering'—the practice of feeding the model foundational documents (style guides, audience personas, brand messaging, and competitive differentiators) before asking it to produce work. By codifying these rules, she creates a 'playground' with guardrails, ensuring the AI output remains aligned with her voice and strategic goals. She notes that the initial effort of setting up these context files is high, but it pays off by allowing for rapid iteration and repurposing of content across multiple formats.
The Last Mile Problem in AI Writing
Parrott argues that AI is inherently limited by its knowledge cutoff and lack of real-world presence. She defines the 'last mile problem' as the gap between commoditized information and unique, high-value insight. Her role as a human writer is to bridge this gap by injecting fresh data, personal experiences, and specific research that the model cannot access. She uses the analogy of a kitchen: the AI is the kitchen and the cooking process, but the writer must supply the fresh, high-quality ingredients to produce a meaningful result.
AI as a Supportive Technology
Beyond professional output, Parrott highlights AI’s role in managing mental health and reducing the friction of 'computer errands.' By automating administrative tasks—such as scheduling medical appointments or triaging an overwhelming email inbox—AI helps her bypass the executive dysfunction often associated with her bipolar disorder. This supportive capacity has fundamentally changed her relationship with work, allowing her to conserve mental energy for complex, creative problem-solving rather than administrative drudgery.
Codifying Workflows with Compound Writing
The conversation introduces the 'Compound Writing' plugin, an experimental tool inspired by Kieran Klassen’s 'Compound Engineering.' The plugin is designed to systematize the process of brainstorming, drafting, and editing. Parrott explains that by borrowing the 'taste' and structural techniques of master writers (like Sorkin or Vonnegut) and encoding them into AI prompts, writers can elevate their editing process, turning AI into a sophisticated collaborator that helps them fall back in love with the page.