Does NotebookLM Create a Good Podcast?

Did creating a podcast just get easier?

Do human beings have to host or produce a show anymore?

What if you just had to feed an AI program the most important information and let it turn into a natural-sounding conversation and call it a day (or an episode)?

Those might be the initial questions when you read or hear about Google's NotebookLM program.

When my wife told me about it last week, I realized I had already heard about something similar from A. Lee Judge on LinkedIn. His example was created by ShowPod.ai.

However, the voices in the AI-generated podcast sound like the voices you get from NotebookLM.

One key difference is that Google promotes it as a program designed to help you summarize in-depth or complex documents by converting them into audio “conversations.”

They call it an "audio overview."

But someone could attempt to turn it into a "podcast episode."

Now, if we look at A. Lee's example, there are some important things to point out:

1. He tells his audience that this is AI-generated. He's not trying to fool anyone.
2. The "show" has some value. It gives users quick insights into the latest marketing news.

I was intrigued.

While I liked what A. Lee did with ShowPod, I found myself pondering other ideas for NotebookLM (other than a podcast).

I wanted to think of a more creative way to use the audio functionality.

My first thought was to create a more engaging audio blog.

Some brands offer an audio version of their blogs and articles in case people don't want to read the whole thing. Instead, they can listen to it.

However, some brands have a generic AI voice that reads the blog, which sounds like the same program Stephen Hawking used to communicate.

OMG - Spend $15 and get someone on Fiverr to read it or something! ??

Anyway...

I chose my recent newsletter about changing SEO strategies to test this all out. I imported the resources/links I used in it as well.

Here's what it sent back:

After listening to the result, I had several reactions to it:

1. The people sound real and very natural.
2. They both say "like," "exactly," and "totally" a lot.
3. They also tend to say things like "Oooo, I like that" a lot.
4. It sounds as if they can't decide if they're knowledgeable people having a discussion or a couple of newbs that are learning new things together.
5. I realized that if this were the shiny new thing in AI, many AI-generated podcasts would sound alike, with the same "hosts" talking on them.

I wanted to take the recording and do two things:

? Edit it like I would any other podcast (there are a lot of extra words and comments that aren't needed).

? Change the voices to see if it makes any difference.

You can already find advice like this on how to edit and change voices.

However, if you're going to do THAT much work, have you made things easier for yourself?

As I told a podcast client last week, people can tell when you put the extra effort into a podcast - and they appreciate it.

However, I'm not done with this.

I'm still going to tinker with the audio a bit to see if there is a way to offer value through audio content.

The AI will likely improve, but that doesn't necessarily make it "better" for you.

As with any AI, there will be improvements in how the voices sound and interact, but that doesn't mean it's a good idea to turn it into a podcast.

Podcasting is so powerful in business because it humanizes brands and builds relationships and trust with audiences.

Content consumers are aware of marketers' and content creators' use of AI, and that's why they will continue to value the efforts put into more humanized content.

Plus, you don't want your audience to feel like you're trying to dupe them like they might if AI had written a book and called it your book.

When I talked with Andy Crestodina on The Content Brief Podcast, he perfectly explained how a human connection will always be part of the book-reader

He also said it's best to view AI as a tool - like any tool you use to help you do something (not take over something entirely for you).

In the end, NotebookLM might be a better tool for something other than creating a podcast.