Warning Our Way To Less Toxic Podcasting

It's hard to have a conversation about using technology to fight podcasting's misinformation and disinformation problems without being called an anti-free speech commie. But I'm going to do it!

Hello, and welcome to another Podcast Pontifications with me, Evo Terra.

Back in May of 2020, Twitter started an experiment in changing the tone, if you will, of the discourse on its service. Now, in brief, Twitter was testing to see if the equivalent of a tap on the shoulder and a subtle *throat clear* could get people to be, well, you know, less of an asshole to one another.

Now, Twitter knows it's very easy to fire off a tweet in anger when you say, or in this case type, things that you may not really mean. And they know that it's easy to get worked up over a headline and blast and retweet or reply without actually reading the linked article. So for the last several months, they've given subtle, gentle, "Are you sure you want to say that?" and "Would you like to maybe read the article before you send your thoughts on it to the world?" messages after someone hits send.

Now, they've announced some results recently, and those results are dramatic.

31% of the people prompted with messages like those changed their behavior. 9% of them deleted their tweet altogether. But 22% revised what they were going to say in the original tweet. Nearly a third of the people, when prompted to reflect on what they were about to do, actually changed their behavior.

Now, granted, we don't have any data on whether or not the changes were better or worse. I understand that. We can hope. Well, you probably know as many people as I know who avoid Twitter because their timeline has become a toxic cesspool. Reducing the toxic cesspool by a third? That's a pretty compelling argument for expanding the test, which Twitter is going to do.

Now, what does any of this have to do with podcasting? Well, I'll tell you. This is a total thought experiment on my part, so set aside your assumptions and come along with me. And this is going to sound like some crazy Big Brother, nanny state, liberal pansy bullshit I'm about to say, but please assume that I'm well aware of that and just play this out with me.

Imagine something like this for podcasting content at the hosting level. You upload your media file and the system goes to work using A.I., Machine Learning, and other advanced algorithms to analyze the words actually said during the episode. And in this fantastical world, potentially problematic sections, let's call them, could be surfaced and brought to the uploader's attention with a message like, "Are you sure you wanted to say this?" with a highlight of the section that's potentially problematic. Maybe even with some resources to help the podcaster understand what might be problematic about the statement.

Now, hold onto your anger and your skepticism because another way to check this could be at the episode details - you still call them show notes, I know - or at the transcript level. This might be a little bit easier because Twitter has already figured out how to parse texts for potentially problematic text. But here again, a gentle, "Hey, you might want to look at this phrasing here," could potentially change some of the podcasters' behaviors.

Now, I know you're in to point out all of the things wrong and dangerous with my idea, and I certainly want to address them. But first, I need to acknowledge that editing a tweet is a much easier prospect than editing a podcast episode that you spent many hours editing and assembling. To quote the noted philosopher, Jules Winfield, it ain't even the same sport.

But what happened earlier in the episode? What if it happened earlier in the episode planning when you could make some changes? Imagine a Yoast or a Grammarly-style plugin that assisted you as you wrote your outline, put your notes together for this particular episode. Or maybe with Descript if you're using that to make a paper edit of your episode before you actually put it into your DAW.

That's not all crazy sounding because we already accept and rely on tools like that to help make our content better as we publish. Again, just mentioned two of them, Yoast and Grammarly. Is it really a bridge too far to have them look out for missteps and, sure, I'll say it, misinformation that we may be unknowingly passing along?

All right. So now's the time when you can yell at me. But before you do so I want to restate, I know how Orwellian this sounds. There are a lot of unanswered questions, like the obvious one who is the arbiter of truth? I also get that this is not going to do anything to curb those who are intentional creating and spreading disinformation on a podcast. I know. I get it. This is just an early idea.

But scrutiny is coming to podcasting. And though it may offend our free speech sensibilities, I, by the way, am a supporter of free speech, still coming. There've been plenty of reports that don't paint podcasting in a flattering light when it comes to disinformation, misinformation, and even hate speech.

And sure, you can assume that doing nothing is what we should do, but I think that approach is equally as naive as what I just put forth. But sure, light me up in the comments.

I shall be back tomorrow with yet another Podcast Pontifications.

Cheers!

Podcast Pontifications is written and narrated by Evo Terra. He's on a mission to make podcasting better. Links to everything mentioned in today's episode are in the notes section of your podcast listening app. A written-to-be-read article based on today's episode is available at podcastpontifications.com where you'll also find a video version and a corrected transcript, both created by Allie Press. Podcast Pontifications is a production of Simpler Media. Find out more at Simpler.Media.