I’ve been a bit obsessed with labyrinths lately. Labyrinths were last month’s topic of a spiritual discussion group I’ve been attending. Since then, I’ve been visiting all the labyrinths in San Francisco.
People have asked me what my meditation process is when I visit one. I’ve adapted a process I found printed near a labyrinth at the California Pacific Medical Center:
- RELEASING: Quieting the mind and letting of the details of daily life as you walk.
- ILLUMINATION: Opening to insight and new awareness when you reach the center.
- INTEGRATION: Moving outward from the center, taking the silence, peace and insight with you into your day.
I’ve adapted this into the more poetic “Introspection – Illumination – Integration” process.
I realized that an infographic might help illustrate this. I decided to see if an AI chatbot could create a decent one for me. So I fired up three — OpenAI’s ChatGPT, Adobe Firefly and Google Gemini / Nano Banana — to see what they could produce. Here’s the prompt I entered for all of them:
Create an infographic about labyrinths. The central image should be a labyrinth. The word “Introspection” should be used to indicate moving toward the center of the labyrinth. In the middle should be the word “Illumination.” The word “Integration” should be used to indicate moving outward from the labyrinth.
Here’s how they did:
ChatGPT
Overly verbose, but honestly I don’t disagree with any of the text. The actual labyrinth is a mess though.

Adobe Firefly
The design is very minimalist and cool, but the content is nonsense. Also, not a labyrinth.

Google Gemini / Nano Banana
And lastly, I turned to Google’s Gemini AI chatbot, Gemini, and it’s media generator “Nano Banana.” This is honestly not terrible. Cool design, very fantastical and dreamy. Clear directions. But of course the labyrinth is an absolute nightmare.

I should say I was using the free versions of all of these tools. I imagine the paid versions would be much more accurate, but I wouldn’t know.
It’s interesting how all of them don’t know how to draw an accurate labyrinth. I assume that is because people often conflate labyrinth and maze, which aren’t the same thing. A labyrinth is supposed to have just one path, with no dead ends or choices along the way.
