In this candid solo episode, Xero opens up about his personal ChatGPT workflow—a behind-the-scenes system he’s never fully shared until now. He explains how he uses ChatGPT not as a crutch, but as a refining mirror to clarify his thoughts for books, characters, and theology/philosophy development. He dives into the importance of maintaining his own voice in the AI responses, describing how he trained ChatGPT by feeding it transcripts from his most passionate podcast episodes to establish a vocal signature.

Xero shares the process of refining ideas through a multi-step back-and-forth with AI, admitting that ChatGPT rarely nails it on the first try. He stresses the need to correct and redirect the AI until it aligns with his ideas—not the other way around. He outlines how he uses multiple AIs for different tasks and how he exports/imports refined threads between platforms to get high-quality results.

But the heart of this episode is frustration: project folders within ChatGPT are disappearing. As of June 8, 2025, Xero has lost two major folders—one of them containing the character dossiers and brainstorms for his entire series. These weren't just final ideas, but the rich, exploratory threads that led to them.

He urges creators who use AI to manually back up their sessions—not just summaries, but the entire dialogue. The process matters more than the outcome, especially if you want to avoid producing what he calls “AI slop.” The episode ends with a powerful reminder: If you don’t know your subject deeply, AI won’t save you—and if you do, you need to protect the process that gets you to greatness.

Timestamps:

* (00:00:03) — Introduction: Why this workflow matters and why he rarely shares it

* (00:00:25) — How he uses ChatGPT to refine ideas

* (00:01:43) — Goal: Make ChatGPT adopt his voice, not the other way around

* (00:02:45) — Feeding transcribed podcasts to teach ChatGPT his tone and intent

* (00:04:00) — Importance of avoiding default AI “bullet point” talk

* (00:05:11) — Using multiple AIs for layered refinement

* (00:06:01) — AI as a research tool—and why Google is no longer reliable

* (00:06:45) — Warning: You must know your subject before asking AI for help

* (00:07:35) — Breaking news: ChatGPT is deleting project folders

* (00:08:53) — His writing method: Web-based brainstorming, not linear drafting

* (00:09:20) — What was lost: character dossiers, wargame notes, narrative setups

* (00:10:55) — Wargaming characters for development and story integration

* (00:11:52) — Providence vs. laziness in backups—lost weeks of work

* (00:12:26) — Why saving full threads helps long-term creative tracking

* (00:13:19) — Call to action: Copy full threads manually, not just final outputs

* (00:14:00) — Final message: Respect the process to avoid AI slop and preserve your art

* (00:14:34) — Sign-off: Peace. Stay holy.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit xeroforhire.substack.com/subscribe