How will AI change the future of podcasts?

As the tools are getting better and new platforms and plugin are popping up, specifically developed for podcast production, it’s safe to say that we haven’t reached peak AI for podcasts yet.

On the post-production side, for challenging audio-recordings, we use Auphonic, iZotope plugins or Adobe Podcast (rarely) ourselves, but they are merely tools at this point and not replacing the skills and years of expertise of the podcaster.

For research and editing ChatGPT is a well embraced king and multiple marketing one-click solutions for TikToks, Reels, jolly posts - or whatever emotion you prefer - and personalised schedules are generated like magic. Integrations like CoPilot in Microsoft 365 and Apple Intelligence lower the bar even further on embracing generative AI.

There’s no question the tools are powerful. The question is how fast the shift in day-to-day production will happen from this in-between state of being tools to blindly trusting an AI with the entire production of your show. Or an essential part of it, like the sound design, video edit and mix of a talk podcast, for example.

Autopod, an AI tool for Premiere Pro for multicam video editing

One of the strengths of experienced podcasters is the trained antennas to sense the emotion of the moment. Take a narrative podcast for example. Delicately put together after months of work, with a carefully crafted copy, voice over and custom music written by a trusted friend to create the perfect symbiosis that takes you entirely into the story. There’s limits on how far AI can be trained. The bottleneck is emotion. The human sensors in creative projects like this can simply not be replaced. Sure, maybe artificial intelligence-tools can analyze a specific feeling in the sound of a voice and then create a stereotypical matching soundscape. But how often have you been enchanted by a podcast that’s checking all the cliche boxes? What about the element of surprise?

Shows that stand out are different. They find the creative angles, surprising approaches, seemingly conflicting elements that work together to create a tickle in your brain or a weird feeling in your gut. Think of a slow motion movie scene where tens of people are brutally fighting, with no audio. You only hear Claire de Lune by Debussy playing at the same time. And yes, Claire The Lune would be a very stereotypical song choice. But you get the point. I can’t see an AI making a carefully crafted decision like that anytime soon.

AI won’t replace the complex skills of the podcast maker. But the tools will for sure get better. We will hear hosts that are fully AI generated and find it hard to distinguish those composed voices from the real ones. Lips in interview videos will be dubbed to correct some info without us even aware of it, as you can already do with short bits of a show in Riverside.fm for example, the online podcast tool.

It’s exciting to see the possibilities coming towards us in a more accessible form than ever. All these little improvements, more likely than not, will be even more integrated into our favorite software. Scripts can already be rewritten to a different tone of voice easier than ever. A one-click-solution for the entire mix of your podcast is already here. With it’s creative and professional flaws, as discusses above, but the race for the developers is ON. I can think of tens of examples of integrations, both online and offline, that will make the work lighter.

What about reaching your audience? In many cases, social media is still the number one way to reach your existing and potential audience. Everyone is fighting for our attention online, the platforms and their algorithms have a creepy amount of power over the effectiveness of it all. How will their power change or shift as their own personalized algorithms evolve, with the power of AI? And will audio-only content ever be decently integrated?

I love reflecting on evolutions like this. There’s so many angles to look at an investigative question like this. How will AI change the future of podcasts? If you want to join the conversation, feel free to reach out! I plan to share interesting new findings regularly. Also tips on podcasts, newsletters, researchers concerning the AI topic are very welcome! niels@wajoo.studio


Niels De Keukelaere

Niels De Keukelaere (LinkedIn) is the founder of Wajoo Podcasts. He started podcasting in 2016 and never stopped. He specializes in podcast optimalisations on the levels of content, sound, branding and marketing, strategy and more and has been working for various companies, media and organisations in Belgium and internationally. Learn more about Wajoo.studio or get in touch!

Previous
Previous

Wajoo webinars and masterclasses, new in 2025!

Next
Next

From the Big Apple