5 Advanced AI Prompts for Deep Podcast Guest Research

The difference between a generic interview and a career-defining conversation comes down to one thing: Research.

If you only read a guest's Twitter bio or their Wikipedia page, you're going to ask the same surface-level questions they've answered on 50 other podcasts. To get a breakthrough interview, you need to dig into their contrarian beliefs, their past failures, and their blind spots.

Thanks to advanced LLMs like Claude 3.5 and GPT-4o, you can do 10 hours of research in 10 minutes. Here are 5 battle-tested prompts to feed your AI before your next interview.

---

1. The "Contrarian Take" Extractor

Most thought leaders have a few core beliefs that go against the grain of their industry. Your goal is to find those beliefs and push on them.

The Prompt:

"I am interviewing [Guest Name], who is known for [Guest's Field/Company]. Analyze their last 2 years of public statements, blogs, and interviews. Identify 3 to 5 highly contrarian opinions they hold that directly contradict standard advice in their industry. For each opinion, provide the context of why they believe it, and suggest one challenging (but respectful) follow-up question I can ask to test their logic."

2. The "Evolution of Thought" Timeline

People change their minds, but they rarely talk about the transition. Asking a guest why they changed their mind is one of the most powerful ways to unlock vulnerability.

The Prompt:

"Act as an expert biographer. I am interviewing [Guest Name]. Track their public opinions on [Specific Topic, e.g., Remote Work / Bootstrapping] over the last 5 years. Highlight any major shifts or contradictions in their stance. How did their opinion evolve? Give me a timeline of these changes and write a question that asks them to reflect on what specific event triggered their change in perspective."

3. The "Blind Spot" Analysis

Every expert has a blind spot—a scenario where their framework breaks down. This prompt helps you find the edge cases of their expertise.

The Prompt:

"I am hosting a podcast with [Guest Name], author of [Book/Framework]. I want to explore the limitations of their philosophy. Give me 3 realistic edge cases or specific industries where their framework would completely fail or backfire. Draft 3 interview questions that challenge them to explain how their philosophy holds up under these extreme conditions."

4. The "Missing Story" Finder

Guests get tired of telling the same founding story. You want to ask about the messy middle—the stories they rarely get to tell.

The Prompt:

"Analyze the public interviews of [Guest Name]. Identify the 3 stories they tell the most frequently (the 'greatest hits'). Then, dig deeper: what is a critical period in their career (a failed launch, a transition period, a quiet year) that they almost never talk about? Formulate a question that forces them to bypass their 'greatest hits' and talk about that undocumented period."

5. The Synthesis Engine

If your guest has written a book, a massive blog, or an annual shareholder letter, you can't read it all in one night. Have the AI synthesize the core mental models.

The Prompt:

"I have attached [Guest Name]'s latest book/manifesto. I am interviewing them tomorrow. Do not give me a generic summary. Instead, extract the 3 foundational mental models or frameworks they use to make decisions. Explain each model in one paragraph, and give me a specific scenario to present to the guest during the interview to watch them apply the model in real-time."

---

How to Execute

Do not read AI-generated questions word-for-word on your podcast. They will sound robotic.

Instead, use these prompts to build a Research Dossier. Read through the AI's analysis 24 hours before you record, internalize the angles, and let those insights guide your natural curiosity during the conversation.

The AI does the heavy lifting of the research. You do the human work of the connection.