Last month I wrote just over 43,000 words. Blog posts, email newsletters, social media content, a few guest posts. That is roughly 10,000 words per week. I did not work more hours than usual. I just changed my writing process.
Here is the system I use. It is not glamorous, but it works.
The Tools
I use three tools in sequence: ChatGPT for drafting, Grammarly for editing, and a simple text editor for final polish. That is it. No expensive subscriptions, no complex workflows.
Step 1: Brain Dump (Monday Morning)
I start the week by dumping every article idea I have into a document. Usually 10-15 topics. I do not filter at this stage. Bad ideas sometimes lead to good ones.
Then I pick the five most promising topics and write a one-paragraph angle for each. What is the point of this article? Who is it for? These two questions eliminate most weak ideas immediately.
Step 2: AI Drafting (Monday-Tuesday)
For each topic, I write a brief outline manually — just headings and bullet points. Then I feed the outline to ChatGPT with a simple instruction: write a first draft in a conversational tone.
I do not expect the draft to be perfect. I expect it to be 70% there. The missing 30% is where the actual writing happens.
Each draft takes about 15 minutes to generate and another 15 minutes to review. Five articles = roughly 2.5 hours of work.
Step 3: Editing (Wednesday)
I run each draft through Grammarly to catch obvious errors. Then I read each one aloud. Reading aloud catches awkward phrasing that silent reading misses.
I rewrite about 30% of each article at this stage. Adding personal examples, adjusting the tone, cutting fluff. This is the most important part of the process. The AI generates raw material. I turn it into something worth reading.
Step 4: Final Polish (Thursday)
I do a final pass on each article, mostly focusing on the opening paragraph and the conclusion. If the opening does not grab me, I rewrite it. If the conclusion does not say something meaningful, I cut it.
Then I publish and move on.
Why This Works
The key insight is that AI handles the volume, but I handle the voice. The system falls apart if I skip the editing step. When I tried to publish AI drafts directly, the quality was acceptable but the articles had no personality. Readers noticed.
By spending 30% of my time on drafting and 70% on editing, I get the best of both worlds: AI efficiency with human quality control.