Episode Summary:
Podcast: The ChatGPT Experiment – Simplifying ChatGPT For Curious Beginners
Episode: 93: The Prompt Master – A ChatGPT Conversation With Keven Ellison
Host: Kerry Weston
Guest: Keven Ellison, VP of Marketing at AIs
Date: December 2, 2025
Episode Overview
In this engaging and practical conversation, host Kerry Weston interviews Keven Ellison, a self-proclaimed “AI Ambassador” and deeply curious AI enthusiast. The episode dives into Ellison’s creation of Prompt Master—a tool designed to help users go beyond basic ChatGPT prompts to create “super prompts” that deliver significantly better results. Together, they discuss best practices for communicating with language models, compare the major AI chatbots, and provide actionable advice and frameworks for beginners and more advanced users alike. The tone remains encouraging, demystifying, and focused on empowering listeners to harness AI with greater confidence—no tech background required.
Key Discussion Points & Insights
1. Keven Ellison’s Background & Motivation
[04:05–06:13]
- Keven explains his 30+ years in tech marketing, evolving from office equipment to the cutting edge of applied AI.
- On why he dove into AI, despite being late-career:
“When they told me that AI was going to be bigger than the Internet, I just went, oh, I have to jump on board on this.” (Keven Ellison, 06:15)
- He sees his role as both an internal advocate—filtering updates and insights for his busy team—and an external resource.
2. The Prompt Gap: Why Most People Don’t Get Great Results
[09:15–12:23]
- Keven observed that 95-99% of people “don’t know how to prompt—not their fault.”
- Common mistake: Treating ChatGPT like Google search; instead, context and specificity matter.
- AI will always deliver something, even if your question isn’t clear, sometimes “hallucinating” or making stuff up.
“It wants to please us... so it does. It’ll make stuff up if we don’t put…guardrails.” (Keven Ellison, 11:29)
- The need for better prompting led Keven to design a scalable solution—Prompt Master.
3. What is Prompt Master & How Does it Work?
[12:51–15:07]
- Prompt Master is an AI tool that ‘interviews’ the user, asking clarifying questions to establish purpose, context, desired outcomes, and any needed resources for the prompt.
- It generates robust, context-aware prompts for ChatGPT, Gemini, Claude, and other major chatbots, recognizing each has different ‘prompting languages.’
- Also serves as an educational tool, revealing to users how much richer their inputs can be.
4. Uncovering the ‘Super Prompt’ & Prompt Engineering Breakthroughs
[17:36–21:35]
- Many viral “prompts”—short and simple—yield surface-level results. The Prompt Master approach builds extended “super prompts” by extracting more detail from users.
- Keven describes using ChatGPT to turn collections of popular prompts into a synthesized, multi-paragraph “super prompt”—a term GPT itself supplied:
“It says, ‘Hey, you want to build a super prompt?’... it goes and builds a three or four page and it has all the questions of what specific things I need.” (Keven Ellison, 20:41)
- Kerry’s framework aligns: What are you doing? Why? What does success look like? Any questions for clarification?
Keven’s Enhanced Prompting Formula
[21:37–24:56]
- Assign a role: e.g., “Expert business writer in the top 0.1%.”
- Give context: Background, specifics, and relevant resources.
- Define success: Output format, tone, examples.
- Encourage clarifying questions:
“Ask me any clarifying questions to ensure you complete this task with 100% accuracy.” (Keven Ellison, 24:26)
- Notable moment: The more consequence or urgency you give (“my job’s on the line!”), the more empathetic and thorough GPT gets (Kerry Weston, 25:55).
5. AI Model Landscape: When to Use Which Tool
[26:00–35:23]
- People experience overwhelm with so many AIs (ChatGPT, Claude, Gemini, Grok, DeepSeek, Kimmy, etc). Keven’s advice: Pick a lane, get the basics right.
- ChatGPT:
- “Best all around product.” (27:11)
- Versions 4o and 5.1 Pro are highlighted.
- Claude (Anthropic):
- Especially strong at making writing “more human”—solid for article and blog rewrites.
- Example: Used Claude to mimic and improve an employee’s style so convincingly that she “was all in” on AI. (30:40)
- Gemini (Google):
- Noted for its image creation (Nano Banana) and Notebook LM tool for infographics and data visualization.
- “You don’t have to do this in Photoshop anymore.” (32:56)
- DeepSeek (China), Grok (Musk):
- Other major players, each evolving rapidly.
- Key takeaway: Tools update weekly; learn foundational prompting, as most tools are just new “front ends” on similar tech.
6. Prompt Master’s Role: Translator for the AI Age
[37:41–37:56]
- Not about picking “the best” tool, but optimizing your approach for each model.
“Think of it as a translator…helping you interact with these AI tools in order for you to get your thoughts out to them, in order to get a great result back.” (Keven Ellison, 37:41)
7. Mindset, Curiosity, and Next Steps
[38:11–39:05]
- Keven is active on LinkedIn (username: Keven Ellison, “with two e’s!”) and shares regular updates and tool links.
- Reiterates curiosity as the core value:
“I love talking about this. I just have a passion for it, and I think that’s what everyone should be because this is our future.” (38:44)
- Kerry’s parting words: “You can’t go wrong being curious. You can’t break it. You can only learn either to do it differently, better, or keep doing what you want.” (38:46)
Notable Quotes & Moments (with Timestamps)
-
On being motivated by massive change:
“When they told me that AI was going to be bigger than the Internet, I just went, oh, I have to jump on board on this.” (Keven Ellison, 06:15) -
On the illusion of “good enough” from AI:
“People…get a response and they don’t know how good their response could be.” (Kerry Weston, 15:14) -
On the Prompt Master’s value:
“I needed to create a tool that would interview me or the person using it, and create a prompt so you can use it in two ways...you can actually learn from it on how much detail you need to put into it.” (Keven Ellison, 13:01–14:07) -
On “unconscious incompetence” with AI:
“We call that unconsciously incompetent…You don’t know.” (Keven Ellison, 15:33–15:40) -
On prompt engineering:
“The last thing is…I ask it, ask me any clarifying questions to ensure you complete this task with 100% accuracy.” (Keven Ellison, 24:23) -
On the power of context and consequence:
“When I actually put those consequences on the line, the questions get better.” (Kerry Weston, 25:55) -
Advice for AI beginners:
“Learn how to prompt first…you don’t need all these other tools. All they’re doing is putting front ends on ChatGPT or on Grok or on Google Gemini.” (Keven Ellison, 37:00)
Key Timestamps for Important Segments
- Keven’s background & AI journey: [04:05–06:13]
- Prompting mistakes & need for Prompt Master: [09:15–12:23]
- Prompt Master explained: [12:51–15:07]
- Prompt frameworks and “super prompts”: [20:05–24:56]
- Model-by-model Chatbot guidance: [27:05–35:23]
- Keven’s LinkedIn & mindset tips: [38:11–39:05]
Resources & Next Steps
- Prompt Master & Image Architect links (as discussed): Will be included in show notes
- Connect with Keven Ellison: LinkedIn – Keven Ellison
- Host’s site for podcast archives & beginner guides: chatgptexperiment.com
Final Takeaway
Whether you’re a curious beginner or looking to deepen your AI skills, this episode offers practical frameworks, real-world examples, and timeless advice: get specific, be curious, and don’t be afraid to experiment. The real secret is not in memorizing “magic prompts,” but in learning how to communicate with AI—and use tools like Prompt Master—to unlock the true potential of these game-changing technologies.
