Do Listeners Dislike AI Voices? Nope.

So here's the recent radio trade headline: "Studies Reveal Growing Listener Disdain For AI Use In Audio." And it's gotta be true because, after all...research.

Here's the key conclusion:

When asked about their favorite podcasts potentially featuring AI voices, 47% of respondents indicated they would be less likely to continue listening, with 28% saying they’d be “much less likely” to tune in. Only 21% expressed increased willingness to engage with AI-enhanced content.

This research was prompted by the splashy debut of a new podcasting company which creates thousands of AI podcasts anchored by AI hosts that are cheap enough to be profitable with an audience of 20 listeners each. Whoo hoo! I used to do magic shows like that when I was 10.

So is this research conclusion actually true?

No, it's not. Let me explain.

First, the question appears to be framed to assess listener response if AI voices were suddenly featured in "their favorite podcasts." So what's being tested here isn't AI. What's being tested here is change. And change to their favorite shows. Well - as you know - the first response of any listener to any change is always negative.

In fact, the real question is how listeners would respond to an AI voice in a brand new podcast, not a favorite one.

And that leads to the next problem...

Listeners are only likely to have an objection to an AI voice if they know the voice is AI. Indeed, the article closes with podcasting guru Tom Webster noting "industry leaders have called for disclosure requirements to protect advertisers and maintain quality standards." Disclosure. In other words, the only way to turn off listeners to AI voices is to signal their presence.

Consider a different kind of research question. Suppose we asked listeners "How would it affect your listening if the local DJ was replaced by a DJ from another city who pretended to be local?" Stand by as the revolt begins, folks. And yet, that is something we do all the time without signaling it.

So, in other words, if you are doing something that is aimed - in whole or in part - at fooling the audience, don't tell them you're doing it and ask them if they like it or not.

And if they only dislike it when it is signaled, then maybe they don't really dislike it after all (relax, all you wonderful voice-trackers).

Needless to say, the remote DJ might be a stronger talent than the homegrown alternative - a better choice for the listener. But that's not what people are evaluating in these questions. What they're evaluating is whether you are trying to fool them or not. And - not surprisingly - nobody likes to be fooled. And a push for transparency is a way to protect talent more than advertisers or quality standards per se.

And there's absolutely nothing wrong with that.

So if you have an on-air gig and you're pondering the future of AI voices on audio, here's my unsolicited professional development advice:

Be great. Be exceptional. Be one-of-a-kind. Expand your skill-set. Wear multiple hats. And do what AI voices can never do: Be present in the real world and touch listeners. Spend more time on the other side of those big studio windows.

Don't ever assume that audiences will revolt at the prospect of AI voices. Because the powers-that-be in the commercial radio and podcasting space are about to unleash an army of invisible voices whether we like it or not. The cost savings will be irresistible in today's sketchy media business climate.

Mark Ramsey Media does audience research for Christian Media - Perceptual research, digital studies, donor studies, underwriter impact studies, music studies, etc. Learn more here. Call Mark at 858-414-4191 or email markramsey@mac.com.

And if you want a strategy to solicit major donors to pay for your research, look here and download this Listener Impact Study solicitation for donors from WAKW-FM.

Want to sign up for this newsletter? Do that here.