The Moat in AI Writing Isn't the Model—It's the Proprietary Data You Feed It
Over two years, writer Katie turned ChatGPT from a career coach during unemployment into a one-person content company and a compounding editor. The core method isn't prompt tricks—it's using proprietary experience to fill the 'last mile' the model can't reach.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
An AI coach's biggest value isn't saving money
After being laid off, she was in a low point and couldn't afford a human career coach's hourly rate, but a $20/month ChatGPT subscription was acceptable. She found that writing out her thoughts and having AI question and push back from the outside helped her externalize thinking, counter catastrophizing, and gave her a sense of accountability. The most important legacy of this experience wasn't the money saved—it was believing that AI could actually drive concrete change in her life. At the time, she was hesitating about whether to take Every's freelance offer; ChatGPT gave her a nudge to just try it.
— Katie ParrottOutput comes from upfront context, not clever prompts
She once took on a delivery load of 8 blog posts, 3 ebooks, 24 LinkedIn posts, 24 X posts, and 16 Instagram posts in a week and a half. She could do it not because she wrote smarter one-line prompts, but because she did a lot of upfront context engineering: feeding the model brand info, product details, audience profiles, and differentiators all at once. Much of the demand was essentially content reuse, and AI is good at rewriting a long article into versions for different platforms. Her rule: set up the scene early, and the rest runs fast.
— Katie ParrottGive the model a fence before talking tone and wording
Before Claude Projects existed, she had to keep persistent documents in Google Docs and manually copy-paste them into the context each time. This structure didn't come out of nowhere—it's the style guide tradition in content marketing: first define who the audience is, what their pain points are, how the product addresses them, and where competitors and differentiators lie. She says this gives the model a fenced playground, ensuring it won't go off track before you even talk about word choice, tone, and reading level. When Projects launched, she set up a separate project for each client and each column.
— Katie ParrottThe last mile of AI writing can only be filled by humans
She believes there's a 'last mile' problem in AI writing: models have knowledge cutoffs and don't exist in the physical world. So the human's job is to supply the latest data and real experience that AI can't reach. She warns that if you just say 'write a blog post about style guides,' you'll get generic content the model already knows—commodity information. What makes writing unique is fresh ingredients the model hasn't seen: internal company research, third-party reports, personal experience. The writing system is the kitchen; the human provides the good ingredients.
— Katie ParrottAI's value is reducing friction, not speeding up output
She speaks openly about her bipolar disorder and points out that AI's value isn't just producing content faster—it's reducing friction in life. She put off booking a primary care physician for three years, then had Codex find a doctor who accepted her insurance and new patients, and complete the appointment. Email is pre-filtered by automation, so only messages that truly need a human reply land in her inbox. She says this completely solved her email anxiety. This isn't using AI as a production tool but as assistive technology that makes her daily operation 'as a person' easier.
— Katie ParrottLet AI interview you repeatedly so it can prioritize for you
Two years ago she simply asked ChatGPT 'can you be my career coach.' Today that coach lives in a Codex project: it contains her job profile, Every's brand positioning, content performance data, a folder of reader praise, and OKRs. She doesn't maintain a dashboard herself; she uses voice conversations to have the system prioritize by highest impact, and the system maintains the Kanban for her. The key to building this system was letting AI interview her repeatedly, then distilling those conversations into context. She says it's like having a Chief of Staff, not just a writer.
— Katie ParrottFeedback that only fixes the current draft is wasted
The core of the Compound Writing plugin is a compounding mindset: every piece of feedback you give AI should flow back into the system so the next output is automatically better. She forked Kieran Klassen's Compound Engineering plugin and adapted it for writing: code and writing both have stages like brainstorming, planning, drafting, and review, except writing focuses more on structure, voice, and evidence. Editing is further split into a substantive edit that looks at big structure and argument, a line edit that handles sentence by sentence, and a final pass for publication checks. Now she doesn't write in a web chat window; she feeds the day's fresh material into this system.
— Katie ParrottVonnegut as editor offers perspective, not verdicts
She built Vonnegut's eight narrative elements, Hitchcock's suspense principle about the bomb under the table, and Sorkin's and Sedaris's expression styles into the plugin as editing skills. Vonnegut checks whether the story starts as close to the end as possible and whether every sentence earns its place; Hitchcock's lens asks whether the reader will want to keep reading. She stresses this isn't about having the master's soul grade your draft—it's giving you a new perspective you can accept or reject. Non-professional writers might benefit even more because they're less likely to nitpick themselves while writing.
— Katie ParrottAI's compounding may only compound for the few who get in early
She says her core thesis about AI today is that education and access will matter more than ever. She lives in Columbus, Ohio, and has never been based in New York or San Francisco, which makes her more attuned to the distribution of opportunity. She admits she got where she is because she had financial slack, time budget, and a network of excellent peers. She hopes the AI community will actively create conditions for more people in more places and of more types to get on board. Because she clearly sees AI's huge compounding effect, the cost is that the gains may only compound for the few who happen to enter early.
— Katie ParrottIn their own words · checked verbatim
Clarity, it turns out, doesn't arrive gift wrapped from a digital assistant or even a human coach. It's something I had to dig out for myself, question by question and prompt by prompt.
Natalia Quintero5:13
And our job as the humans is to close that last mile and provide the real world experiences that AI can't get because they come after the knowledge cut off and it's also out there happening somewhere in the real world and AI is not in the real world, the physical world yet.
Katie Parrott15:40
We think about AI as a productive technology, right? We value it for its ability to produce output and get things out of our, like just put stuff out. And what I found in my personal experience is that AI is just as powerful as a supportive technology.
Katie Parrott22:50
compounding is the idea that every piece of feedback that you give to AI should feed back into the system to improve the next output.
Katie Parrott32:27
My thesis on AI today is that education and access are going to matter more than they ever have before.
Katie Parrott44:49
But the risk there is that value is going to compound to the small subset of people that happen to be early. And I just think that a much more compelling vision for an AI future is one where everyone is able to come along for the ride.
Katie Parrott46:32
Figures
| Human career coach hourly rate vs. ChatGPT monthly fee | about $150/hour vs. $20/month | 3:11 |
| Time she put off booking a primary care physician | about 3 years | 23:53 |
| Security vulnerabilities Codex found in her Tastemaker project | 5 | 43:24 |
Glossary
- context engineering
- Pre-organizing brand, audience, product, and differentiators into a fixed background so every AI conversation heads in the right direction.
- last mile problem
- Models have knowledge cutoffs; humans must supply the latest data and real-world experience or the writing falls into generic clichés.
- compound engineering
- Sinking one-off feedback to AI into the system so the same good decisions take effect automatically and repeatedly in the future.
- computer errands
- Having AI agents handle administrative chores you keep putting off, like booking a doctor, filing insurance, and triaging your inbox.
How to listen
Independent writers and content entrepreneurs frustrated by the quality of AI-generated drafts, plus anyone who wants AI agents to take over life and administrative chores.
The opening minutes about getting laid off and the ad breaks can be fast-forwarded; start at 8:26 with the delivery case study to get to the methodology.