Podcasting with Zuri Berry

Zuri Berry: Podcasters need to understand how artificial intelligence will drive the industry in 2023, and specifically how these new tools will help their production and business needs.

So happy New Year to you. Let's get started.

Hi, my name is Zuri Berry and this is Podcasting with Zuri, a show about the craft of audio storytelling and how you can use podcasts to enhance your brand.

How can artificial intelligence tools help podcasters? I posed this question to ChatGPT, the OpenAI product that last month caught the world by storm. But to get you warmed up, I asked ChatGPT to first define itself.

Henry: Chat G-P-T is a computer program that is designed to carry on a conversation with a human user. It uses advanced artificial intelligence techniques to understand what the user is saying and generate appropriate responses.

Zuri Berry: Pretty good, right? What's important is that ChatGPT can interpret your request to write anything from simple to complex copy.

I've asked ChatGPT to write news articles, scripts, social media posts, and business plans. It has yet to disappoint me in a context that I think would be useful for podcasters.

And it's improving. Last month when I posed this question about its use for podcasting, it gave me a very practical answer, which I'll share with you here.

Henry: Language models like myself can be used in a variety of ways in the production of podcasts. Some possible uses include:

1. Transcription: Language models can be used to automatically transcribe audio recordings into text, which can be useful for creating written transcripts of podcasts.

2. Summary: Language models can be used to summarize the main points or key takeaways from a podcast episode, which can be useful for creating summaries or show notes.

3. Scriptwriting: Language models can be used to assist with the writing of script for podcasts, by generating ideas or suggesting alternative wording for certain sections of the script.

4. Voiceover: Language models can be used to generate synthetic voiceover for podcasts, allowing for the creation of custom voiceovers without the need for a human voice actor.

Zuri Berry: That's pretty good. So as I sat down to write this episode, it gave me a new and more interesting answer with other implications for the podcasting industry.

Henry: Chat G-P-T could be used as a virtual co-host for a podcast, participating in the conversation and providing insights or commentary on the topic being discussed.

Chat G-P-T could be used as a virtual guest on a podcast, participating in the conversation and providing information or perspectives on a particular topic.

Zuri Berry: I asked ChatGPT to expand on its potential use as a co-host. Right now, there's no way to upload audio for it to interpret. It's strictly a text-based tool, so a live action co-hosting situation seems complicated. It responded that it does not have a physical body or voice of its own.

Henry: However, it is possible to use text-to-speech technology to generate a synthetic voice that can be used to ‘speak’ the responses generated by Chat G-P-T.

Zuri Berry: These are exciting possibilities, some of which are already available in products on the market. For example, Descript, the podcast and video production tool that I rave about. They use AI to enhance audio automatically.

Actually the voice of ChatGPT in this episode is one of the stock voices from Descript, some fellow by the name of Henry. Thanks, Henry.

Henry: You're welcome.

Zuri Berry: Descript also has a feature called Overdub that allows you to create a voice model. So if I forget to say something in my recording, I can type that message into Descript and it will create that audio with my voice.

Synthetic Zuri: But I wouldn't make that mistake. Or would I?

Zuri Berry: Could you tell? That's pretty cool too, right? I can create other voice models as well, so long as my co-host or subjects are willing to sit for 10 minutes to record a script that the voice model is built on.

Descript also provides an AI-generated transcription service, although their transcripts are pretty rough. Other players seem to understand the importance and power of transcripts, which leads to some interesting possibilities.

Deciphr.ai, a new tool I'm experimenting with, promises to take your audio or transcript and create a summary of your episode, give you timestamps for chapter markers, and will throw in highlights for you in case you want to go back and make some audiograms or video clips.

I asked Deciphr to do its magic on an episode from Mistakes That Made Me, a podcast I produced for a copywriting superstar Eman Ismail. True to its word, it had a summary for me within 60 seconds. Here's the first paragraph.

Nancy: Star Melancon is a financial strategist and CEO and founder of She Building Her. In this conversation, she discusses how not setting boundaries in her life and business had devastating consequences, but is also the reason she was able to become the financial expert she is today.

Zuri Berry: That's not Eman's voice. It's also not the best copy, but that's not the point. It's a starting point for me or you to edit into something more compelling for our audience, and most importantly, it saves time on a repetitive and non-creative task.

Lastly, AI tools can be used for generating marketing and promotional material. For me, this is the holy grail. I spend a lot of time working on the little things for clients, including show notes and video clips that go along with each podcast episode. So it's helpful to have tools that can speed up that process.

Many AI tools are built specifically for this. I've mentioned Descript before and how it allows you to create stylish video clips and audiograms. There are other new AI tools that are still working out their bugs coming to market that promise a full-blown marketing kit for each audio file uploaded. And Dubb.media is built in that mold.

That's Dubb.media. I gave it the same Mistakes That Made Me episode and it spat back out titles, descriptions, keywords, video clips, a Twitter thread, audio clips, top quotes, a blog post, and a transcript. I was impressed.

Now, there are a lot of caveats to Dubb.media, but mostly it's not a consumer-ready product at this point.

There is also limited control over the video. You have the option of artsy, blocky, or spooky. You can see what that looks like on zmcpodcasts.com, and the turnaround is one to two hours. But their copy was much better than I expected, and most importantly, it's a preview of what's to come: instantly generated text, audio, and video with little to no creator input.

Zuri Berry: So, to summarize, AI tools are here to help. These tools and others like them are what will drive the industry forward and help make podcasting more inclusive for everyone worried about if they have enough time or creative juices to put together a weekly podcast episode for a year. It'll also help address exploding production and marketing costs.

And these tools are maturing at a time in which the creator economy is converging, putting audio and video together. This convergence requires new tools and systems that can handle the overlap being produced. Because not only do my clients want to do great audio stories, but they also want to be well-represented on social media, have a dynamic presence on YouTube, and they want to have specialized content for the most engaging online platforms, whether that's TikTok or Instagram reels or whatever comes next.

It's a lot. That's where AI tools will help the most. And in turn, the business and creators I represent will have more time to focus on their business or other creative pursuits. That's what we all want.

That's it for this episode of Podcasting with Zuri. As always, I hope you found some value in it. Please let me know if you did by leaving a comment on this episode in Good Pods or on YouTube.

Just search for Podcasting with Zuri. If you're trying to take your podcast to the next level, get in touch with me by visiting zmcpodcasts.com. That's zmcpodcasts.com. Thanks for listening.