It’s not an exaggeration to say that artificial intelligence exploded onto the podcasting scene this year.
Every week, it seems there’s a new AI-powered podcasting tool that claims it’s about to totally revolutionise how we create, distribute and promote our podcasts.
With this technology evolving at a shocking pace, there’s no denying that the future of the industry is going to look quite different to what we know now.
AI will make podcasting more regulated
At The Podcast Show this year, Bryan Barletta from Sounds Profitable made an interesting remark about the future of the industry. Like it or not, he said, podcasting is about to get a lot more regulated.
He links this prediction to the industry’s economics, which failed to hit its predicted $2 billion value last year. According to Barletta, some of the biggest advertisers still see podcasting as too unregulated to invest in. They can’t risk being affiliated with podcasters who say whatever they want without any thought about the impact of their words. But as we know, some of the most popular hosts in our industry are also some of the most controversial.
It looks like AI is going to play a leading role in this move towards regulating the podcasting industry. Barometer has already partnered with NewsGuard to develop an AI-powered tool that can detect misinformation in podcasts. They also have a separate tool that can assess brand safety and suitability at an episode level.
These tools can scan through masses of audio content using a tagging system and machine learning algorithms. The technology can then flag any episodes that discuss potentially controversial topics or include swearing so that advertisers can avoid them.
Holding podcasters accountable for spreading misinformation or hateful narratives is absolutely a good thing. But there’s no denying that part of what makes podcasting so special – and why listeners love it so much – is its unfiltered nature.
The idea that podcasters will avoid discussing controversial topics or using curse words just because it might impact advertising opportunities feels very contra what the medium is all about. But we’re already seeing this happening. Even Ben Shapiro – notorious for his ‘loose’ speech – is censoring curse words from his shows so these tools don’t flag his podcast as unsafe for brands to work with.
AI will dismantle language barriers
Natural Language Processing (NLP) is a branch of AI that tools like Descript and Whisper use to automatically generate transcripts. NLP is just one example of how AI eliminates some of the most time-consuming manual tasks we must handle as creators.
Soon enough, however, we’re likely to see podcasters using NPL to break down language barriers for the industry too. And I don’t just mean making podcast transcripts available in other languages. I mean audio too.
With a mixture of NLP and voice-cloning technology, we’ll likely start to see every podcast show available in multiple languages. Think about the process: you create a transcript of a podcast episode using NPL and translate it into another language in text (a Google Translate integration would easily handle this). Then, a combination of text-to-speech and voice-cloning technology could auto-translate your podcasts into multiple other languages in your own voice.
Just think about what this could do to help you break into new international markets with your podcast. Within minutes, your podcast could be available in Spanish, Portuguese or even Chinese Mandarin for the very first time. It’s mind-blowing stuff.
New tool launched to help businesses kickstart their podcast
AI will be exploited to sell more stuff
While voice-cloning technology clearly has its perks for podcasters, it won’t enter the industry without at least some controversy.
A while ago, we learned that AI radio DJ software is being sold to radio stations so they can generate localised ad reads in the voices of existing hosts. A great resource-saver for struggling local radio stations. What could possibly go wrong?
Well, the problem is that when you use voice-cloning to sell stuff, it undermines the trust bond between listeners and hosts. We know that podcasts are the most effective medium for advertising (more effective now than both TV and radio). This is why brands are so keen to work with us.
We also know that host-read ads are the most impactful because listeners trust their favourite podcast hosts a lot. While using voiceover technology in this way isn’t technically exploiting the host-listener bond, it could reduce the impact of podcast advertising over time as listeners become less trusting of the messages they hear.
Spotify is apparently already doing this. In a recent episode of his podcast, Bill Simmons told Derek Thompson of The Atlantic: “I don’t think Spotify is going to get mad at me for this, but we’re developing that stuff… There is going to be a way to use my voice for the ads.”
And what is there to stop advertisers from cloning the voices of trusted hosts to sell products even without a host’s consent?
AI will elevate the value of human-generated content
AI reduces production barriers, making it easier than ever to create podcasts, so it’s inevitable we’ll see many more new shows popping up soon. But quality content? That’s a different matter.
As barrier-to-entry hits the floor, newbie creators will attempt to enter the market with content that’s entirely AI-generated from start to finish. Some will use text-to-speech technology so they don’t need to go near a microphone to produce a podcast. But just as we saw during the pandemic, creators who jump into podcasting without the right level of passion and commitment don’t last long.
Those creators who focus on keeping their content personal and human – only using AI to speed up manual, menial processes like creating transcripts – will reap the benefits. As the world of content becomes more saturated with AI-generated material (and this goes for written content too, not just podcasts), this will only drive up the value of unmistakably human-generated content.
Again, the bond between podcast hosts and their listeners is unique and vital to podcasting. This is why computer-generated content just won’t catch on in this industry. We’ll likely see a lot of AI-generated podcasts appear and disappear just as quickly.
Katie Paterson is a content feature writer at The Podcast Host, maker of Alitu – a simple tool for recording, editing and publishing podcasts