Creating a Podcast with NotebookLM; What are the negatives and positives of that process?

I first heard about NotebookLM from Cristina Capretta in the pre-conversation on my podcast and, to be honest, never really thought much about it. She shared that this new technology could take text or something else and create a conversational podcast between two people that sounded pretty authentic. I probably didn’t pay much attention to it because it seemed too surreal.

Then I saw it being shared on social media and thought I would try it myself.

I uploaded a PDF of a recent email in which I wrote about some of my takeaways from Adam Grant’s latest book, “Hidden Potential.” With a few clicks, a conversational podcast was recorded, and the results blew me away.

Was it perfect? Absolutely not.

I noticed some misattributions in the podcasts right away, and I am not sure if it was because of my writing or for a different reason.

But it was also not perfect, which also was concerning.

The flaws made it feel somewhat authentic, which makes the technology even scarier, in my opinion. How easy would it be to manipulate a conversation when it genuinely seems like two people are having a conversation?

Although it only records a conversation between a man and a woman in the English language, technology today is currently the worst it will be moving forward.

Here is the video (you can also listen to it on Apple Podcasts and Spotify), where I introduce the podcast and share some thoughts before I share the AI conversation with others:

I did some “App Smashing” with this process and did the following:

1. Uploaded a document to NotebookLM.

2. I made a visual video using Pictory.AI (the free version).

3. I took that video and uploaded the audio to CapCut, where I auto-generated captions (using AI) and did some editing.

The whole process took about 30 minutes to create. Since it was my first time, I know I will get faster.

I wanted to share the podcast not because I think it is amazing but to understand how to utilize it from a learner’s viewpoint and hopefully spark some conversation in school communities.

When ChatGPT first came out, I posed some questions to consider before considering the implications of school usage, and they can be easily adapted for this process.

1. What are the negatives of this technology, and what are some of the potential benefits?

2. How could you use this technology in your personal life?

3. How could this improve opportunities in your professional work?

4. How could this be utilized in student learning?

These questions are in a particular order for a reason.

1. What are the negative and positive aspects of this (in that order)? Let’s identify potential obstacles and then see if there are positive opportunities.

2. Questions two and three have a connection. Do I see the value of this in my personal and professional life? If I don’t, why would I ever teach it to students? If I do, how could I not teach it to students? We are more likely to share and teach something to students if we see its value in our own lives.

3. Then, through your own learning process, do we consider how we learn it with students?

Too often, we want to jump to question number four (How could this be utilized in student learning?) without thinking about our own experiences. By creating this podcast and sharing it with others, I am going through this process myself.

Interestingly enough, as soon as I finished creating the video, I saw this TikTok about how AI podcasts suck because we feel a connection to the people making the podcast, not just the content.

@thedylanschmidt AI podcasts suck. #ai #podcasts #podcasting #artificialintelligence ♬ original sound – Dylan Schmidt

There is definitely some truth to that.

In the podcasts I often listen to, I feel I know the person or have some connection. The rawness of it and the human connection matter. Maybe AI will learn to fake that as well.

My friend, AJ Juliani, often shares the following (paraphrased). I am not pro-AI or anti-AI (insert any tech), but I am pro-learning.

I hope that by sharing my own learning, you will figure out your own way and consider how we can use this technology to not only make our lives easier but also make us think deeper, become more curious, and connect to each other on a more human level. Does utilizing this type of technology do the opposite?

I will leave that question to you, or we can wait for our AI Overlords to decide for us. I think I will choose the former.

P.S. This post was written by me (probably poorly), and I utilized Grammarly to make it easier for you to read.