Description

Podcast Notes

Episode 483 The Promise and Pitfalls of ChatGPT
Hosts: Brian Miller, PCC and Chad Hall, MCC
Date: September 18th, 2025

In this episode, the hosts reflect on their early experiences experimenting with ChatGPT, exploring both its surprising strengths and its frustrating limitations. What began with hesitation and even skepticism turned into curiosity as they discovered that working with AI requires learning a "new language." They share examples of using ChatGPT for brainstorming, writing support, and planning, while also wrestling with issues of trust when the tool generated confident but false information. 

Key Highlights

• Early reluctance to adopt AI gave way to curiosity and intentional use, framed as learning a "new language."

• ChatGPT was tested as a professional thinking partner, helping with brainstorming, writing support, and planning.

• Trust issues emerged when AI confidently gave false or misleading information (e.g., fabricated movie scenes, podcasts, or emails).

• Iterative, back-and-forth prompting produced better results than one-shot requests.

• Teaching the model to mirror personal writing style improved authenticity and usefulness in content creation.

Takeaways

• Treat AI as a collaborative partner, not a one-click solution.

• Be aware of AI's tendency to "hallucinate" information and verify claims independently.

• Strong prompts and iterative dialogue are key to getting useful, authentic-sounding results.

• AI can accelerate creativity and productivity, but final ownership of content must remain human.