VENDLER.

Experiments

2025.03.17.

I've always been curious about how much of my security I'm willing to risk for my freedom. Or the other way around.

I experiment with my boundaries, trying things I have no prior experience with. One such experiment was my "pay what it's worth" business model. I think it's a great idea, but to this day, no one has tried it. (Be the first!)

So I made a change.

A few days ago, I shared a business offer that allows company leaders or aspiring entrepreneurs to pay for my consultations either with a fixed amount or by contributing to the conversation itself—through their insights and shared experiences.

"Pay or Share." And... it's working!

So far, in the past five workdays, I’ve received nine inquiries and three podcast invitations. From these inquiries, I’ve scheduled two online meetings and seven dog-walking consultations. Out of the nine people, one expressed interest in paying, while the others gladly contributed by sharing their stories and experiences.

I couldn’t record the first conversation—I don’t yet have the proper setup—but by the end of the day, I had two recordings. I also started experimenting with how AI could help turn these into useful case studies.

First experiment:

  • I uploaded a 60-minute audio recording to TurboScribe, which transcribed the Hungarian text in about 10 minutes.
  • I then imported the text into NotebookLM, which processed it in a minute, generating a summary and an excerpt.
  • Using NotebookLM, I created a 22-minute English-language podcast conversation in about 10 minutes, then transcribed it with TurboScribe, hoping to create a more engaging and concise source (but I didn’t).
  • I uploaded the content to ChatGPT and asked it to generate a case study.

Observations:

It’s clear that NotebookLM is incredibly good at summarizing. You can ask it questions, and its responses are well-structured, logical, and precise. ChatGPT can already produce a coherent and correct draft, even at length, but it struggles with emphasis, requiring significant post-editing.

I can imagine that a publicly accessible AI system, where I upload raw case study materials and allow users to freely ask questions, might be a better approach. It could be more useful and easier to navigate than a traditional collection of case studies.

Such a system would contain the stories, dilemmas, and life situations of those who reach out to me—along with the recurring discussions about these topics. A wealth of experiences, emotions, and decisions...

I’ll give it a try. And I’ll share it with you.

In the meantime, if you have any ideas on what I should experiment with, or any thoughts, suggestions, or critiques... let me know.

--

The article was translated from Hungarian to English by ChatGPT. Thank you, ChatGPT, for being here.

2025. BALAZS VENDLER

All rights reserved.