As a Substack content creator, finding fresh and engaging ideas can often feel challenging. ChatGPT and other free AI tools have become popular aids in the creative process, but I’ve discovered that relying on them completely is not the solution. Instead, I treat these tools as valuable collaborators that work alongside my own creativity and a carefully maintained idea repository.
In this article, I’ll share my exact approach to brainstorming with AI so you can tailor it to your own style and workflow.
Maintaining an Idea Repository
Before I even open ChatGPT, I keep a dedicated idea repository—a simple note or document where I capture every spark of inspiration, topic suggestion, or passing curiosity. This practice ensures I never face a blank page and always have a rich pool of ideas ready to explore.
When it’s time to create new content, I review my list and pick one promising idea to develop further. This approach has saved me countless hours and made the creative process much smoother.
How I Use Large Language Models (LLMs) for Brainstorming
Step 1: Start with a Clear, Detailed Prompt
The key to getting useful output from ChatGPT or any LLM is crafting a clear and detailed prompt. Define a specific role for the AI, provide relevant context, and explain exactly what you want.
For example, a prompt I use often is:
“You are a creative writing assistant helping me brainstorm blog post ideas about self-publishing. Please suggest five unique angles with brief explanations.”
The more precise your instructions, the better the AI’s responses will be.
Step 2: Feed Relevant Content to Teach Your Style
To make AI suggestions feel authentic and match your voice, provide samples of your past writing—newsletters, articles, or even social media posts. This helps the AI understand your style and think like you.
For example, just last week I pasted a few paragraphs of my own writing into ChatGPT before brainstorming a new post about creator burnout. The suggestions that followed were immediately more on-brand compared to starting with a blank slate.
Step 3: Keep the Interaction Conversational
Brainstorming with AI is not a one-time task. Treat it as an ongoing conversation. Build on ideas, ask clarifying questions, and explore new directions.
Some of my best topics emerged after several rounds of back-and-forth. Think of AI as a collaborator that challenges and expands your thinking, not just a machine that delivers finished work.
Bonus Tip: Cross-Validate Ideas Using Multiple LLMs
If you have access to more than one AI tool, try taking ideas or drafts generated by one and feeding them into another, such as Claude, Gemini, or Perplexity, for additional feedback. This cross-validation often sharpens your concepts and reveals angles you might never have considered.
Final Note
Large language models like ChatGPT can be powerful partners in brainstorming, but they are just partners. They cannot replace your creativity, judgment, or unique voice.
When you learn to use these tools effectively, you’ll unlock new ideas faster and more consistently than ever before.
Ready to get started?
I’ve created a free PDF with my favourite example prompts to help you brainstorm more quickly and easily.
👉 Supercharge Your Idea Generation with AI
Or simply reply to this message and tell me what you’re working on. I’d love to help you shape your next great idea.
Weekly Insight
I share one writing tip like this and one original short story each week, both crafted to help you grow as a writer and a reader. Stay tuned for more tools, techniques, and inspiration.