top of page

The Systemic World Podcast: How AI is Powering My New Experiment

Generative AI at Unsplash.com
Generative AI at Unsplash.com

Lately, I’ve been experimenting with AI to create an audio-based extension of my work in systems thinking and storytelling. Enter The Systemic World Podcast—a fully AI-generated audio experience that brings complex stories to life. This project emerged from a simple curiosity: What happens when we let AI analyze, reinterpret, and vocalize the systems hidden within our narratives?

The Process: From Articles to AI-Generated Conversations

The workflow is relatively simple but involves several key steps to fine-tune the final output. Here’s how I do it:

  1. Uploading Articles to NotebookLM I take articles previously featured in the Systems and Stories newsletter and upload them to Google’s NotebookLM. This AI-powered research assistant processes the text and generates a conversational discussion between two virtual speakers.

  2. Optimizing the Configuration Prompt To ensure the generated conversation stays on point, I tweak the Configuration Prompt in NotebookLM. This allows me to include specific instructions such as:

    • A welcome message to introduce listeners to the podcast.

    • Emphasis on key takeaways from the article(s).

    • A natural flow in the dialogue to make it feel like an insightful conversation rather than a robotic recitation.

  3. Reviewing and Refining the AI-Generated Audio Once the conversation is generated, I listen to the output, checking for coherence, clarity, and engagement. AI-generated content often introduces unexpected nuances—sometimes making insightful connections, sometimes going off-track. I refine the responses where necessary to enhance the storytelling.

  4. Enhancing the Experience in Canva After finalizing the audio, I import it into Canva. Here, I create a simple but effective visual representation—essentially a podcast “cover” that acts as the backdrop for the conversation. This turns the podcast into a video format, making it accessible for platforms that require a visual element. where it reaches a wider audience and serves as a visually engaging extension of the Systems and Stories narratives.

  5. The podcast is then uploaded to YouTube, where it reaches a wider audience and serves as a visually engaging extension of the Systems and Stories narratives. You can listen to it here: The Systemic World Podcast.

The Systemic World Podcast Cover.
The Systemic World Podcast Cover.

AI as a Creative Partner

What has been really fun about this experiment is observing the biases that AI introduces and how it understands the basics of mental models and systemic interactions. While it’s far from perfect, it’s a fascinating starting point. AI doesn’t just replicate content—it interprets, reframes, and sometimes even enhances the ideas within the original articles. This introduces an additional layer of complexity, as the generative AI shifts and adapts based on the triggers within the story, making the conversation feel dynamic and evolving.

So far, I’ve tested this process with seven articles from Systems and Stories, mainly for fun and to explore the potential of AI in storytelling. What I’ve discovered is that AI doesn’t just transcribe; it interprets. It sometimes introduces additional layers of complexity, providing clearer examples or reframing ideas in ways that I wouldn’t have considered on my own.

This ability to reinterpret and expand upon a narrative adds an intriguing dynamic to the podcast. Instead of merely echoing the articles, AI acts as a collaborator, offering fresh perspectives that may challenge or refine the original message. In a way, it’s like having an AI-powered co-host that continuously finds new angles within the story.

What’s Next?

Right now, The Systemic World Podcast is just an experiment—one that started as a playful way to extend my work in systems thinking. But as I refine the process and gather feedback, I’m excited to see where this could lead. Could AI-powered podcasts be a new way to engage audiences with complex topics? Could this approach enhance education, storytelling, or policymaking?

For now, I’ll keep exploring, iterating, and sharing my findings. If you’re curious about AI-generated storytelling or have ideas for topics, let’s discuss! The systemic world is full of surprises, and AI might just help us uncover a few more.

コメント


bottom of page