A couple of months ago I had ChatGPT write a post about hiking the Camino imitating my style. I thought it was pretty remarkable. Now I’ve been taken to a new level of amazement at what these large language models can do.
Yesterday at church was our annual Messiah sing-along and potluck (those two things are not odd bedfellows in progressive Mennonite culture). Seated at my table for the lunch was the tenor soloist, Nate, who is also a software developer (also not a strange pairing in our community). Our conversation went to AI and he told me about the Google AI “Notebook LM” which can take a document and turn it into a podcast. I said, “you mean it just reads the document to you?” which I thought to be less than remarkable these days. “No, it creates a podcast conversation between two people who explain the content of the document podcast-style.” Wow, I thought, that would be remarkable.
So today I went to the website and gave it the pdf of my book, The Sacred Chain and instructed it to make a podcast about it. It took about 3 minutes to generate these 14 minutes of conversation, which is really really remarkable.
The inflection on the voices is incredibly realistic, and its “understanding” of the book is unbelievable. It picked out the salient points of each of the five parts of the book and had some bantering discussion about it. The only obvious mistakes were that it said “four parts of the book” when there are five (though it discussed all five), and that it mistook the parable I told in the last section for something slightly more literal and realistic.
Give it a listen. Perhaps it will entice you to get a copy of the book and read it for yourself before books too have become obsolete. I fear that AI and the art of creating books and podcasts will increasingly be an odd combination in our culture. Let’s make the most of it while we can!
And for bonus AI content, here’s the image it made to go with this post (which you can only kind of see in the background of the recording above):
Share this post