Mind Readings: The Real Danger to the News Industry Isn't AI

In today’s episode, I delve into the looming crisis for the news industry: will AI be the final nail in the coffin? It’s not AI generated content that’s threatening the industry, rather, the danger lies in the fluff-filled articles that bury the actual news beneath paragraphs of filler. AI, especially models like GPT-4, can distill these lengthy articles, extracting the crux while leaving behind the fluff. This potential evolution might significantly impact advertising revenues, given that AI won’t click on ads, lowering their effectiveness. But, is it all doom and gloom? Maybe not, if we adapt. I discuss how platforms like Substack are creating new revenue models for content creators, where direct communication and interaction with the audience is prioritized. Tune in to understand how the future of content creation and publishing might need to evolve, and why it’s vital for you to create valuable content that holds the reader’s interest, rather than fillers. Don’t forget to hit that subscribe button if you enjoy these insights!

Summary generated by AI.

Mind Readings: The Real Danger to the News Industry Isn't AI

Watch this video on YouTube.

Can’t see anything? Watch it on YouTube here.

Listen to the audio here:

Download the MP3 audio here.

Machine-Generated Transcript

What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for watching the video.

In today’s episode, let’s talk about the news.

And the danger to the news industry that AI might or might not.

Cause a lot of people are talking about how the news industry is on death’s door.

And AI is going to be the final nail in the coffin to put newspapers and other publications out of business.

Maybe, but probably not.

What’s going to kill the news industry is the garbage that they publish.

I was recently looking for click looking for some some news about one of the the shows that I follow.

I don’t have time to watch television, but I read the summaries.

And I got to this one site that has interesting headline, and just scrolling, scrolling, scrolling, okay, when we get to the actual news that, you know, the headline said one thing, and then there’s like 14 paragraphs of filler, just total fluff, saying nothing.

And then finally, at the very bottom, the last paragraph is the actual news piece.

That’s a miserable experience.

Now, why would a news website do that? Well, because to scroll past all 14 paragraphs, if you do that, in a normal web browser, one that does not have an ad blocker.

There’s like an ad, every paragraph, so they’re just showing ad after ad after ad, as you’re trying to scroll through the thing, you know, just being boggled by the lack of content.

I would have loved to say that that was all AI generated.

But it wasn’t it was just badly written human content, actually did a test with one of the many AI detection tools.

And they all universally agreed.

The site’s not the you know, the content here is not AI written it’s it doesn’t have the telltale statistical indicators that hey, I generated content, which by the way, they do exist.

It’s a technique called perplexity and we’ll talk about that another time.

But holy crap, that article was so long and so drawn out for so little benefit that it was just garbage.

It was all filler, no meat.

Now, when I fed that article to OpenAI to GPT-4, I said, summarize this, and just give me the main points, and it did it it went straight to the main point, cut out a lot of the filler.

And that was a huge time saver, that technique is a huge time saver for like, Oh, my goodness, just piles of dreck.

machines like that, and large language models and AI have the ability to summarize, to distill to extract to remove information from whatever soup It’s in and boil it down to just the relevant parts.

In fact, in terms of what large language models are like, like a ChatGPT, based model GPT-4 were llama or any of these other ones.

They’re really good at that they’re really good at summarization and extraction, they’re actually better at that than they aren’t generation, that they’re better at extracting and summarizing than they are at writing net new content.

And that’s one of the great uses of these tools.

It is fairly trivial to envision software that you would have as an app on your phone, whatever that goes and reads all these poorly written news sites and just says here’s the two bullet points from this article that are that are relevant.

And the rest, you know, we’ve we’ve ignored because it’s all filler, it’s all fluff.

That’s what’s going to kill the news industry.

That’s what’s going to kill a lot of journalism, it is not machines, putting writers out of work.

It is machines, distilling down the garbage that’s being written, and in the process, not driving ad revenue, right, because a machine that goes and parses that page, it’s not a human, right, it’s not, it is running a browser.

So the publisher might get some views on those pages if it renders it in a contained environment.

But they’re not going to get clicks on it ever, the ad performance is going to drop to zero because a machine is not going to click on those ads and machine is instead just going to take the text from the page, boil it down to the one bullet point that is actually the news.

And there we have it.

So that’s a pretty bleak picture.

If you’re a publisher, right? Machines are going to be reading your content and distilling down just the bits that people want and leaving the rest behind and you’re not going to get any clicks.

So you may get ad revenue, but you will not be the advertisers will be like it’s this is not paying off.

We’re advertising we’re spending money.

And we’re getting no results.

We’re getting no traffic on these ads.

So what’s the solution? Well, there’s two solutions one, create less crap.

And to the model for how publications do business has got to change and and what it might look like is what is being very successfully done now on places like substack, where you have individual writers creating their own feeds of things.

And then having sponsors, right? Have a, I can’t tell you the number of newsletters I read now that have a sponsor, and yeah, you read it.

And ad blockers don’t cut it out.

Because it’s an email.

It’s an email, and you just scroll past the ad, if you’re not if you don’t care.

But if you do care, the ads right there, and you can read through it, and enjoy it.

I look at my friend and handle these newsletters.

She’s got ads in it for some of her stuff.

I look at something like, what’s that guy wrote in his AI rundown newsletter, I can’t remember his last name.

He’s got promotional stuff in his newsletter, all these different newsletters that people are subscribing to now, that trend is taken off because A, it allows writers to talk directly to their audience without the constraints imposed by a publisher, and B, they can make money directly from the audience by charging for subscriptions, in some cases, by running sponsors, things like that.

That’s the model for publishing that seems to be working right now.

People who are good content creators are creating their own publications, their own platforms.

And in doing so they are able to derive revenue from it.

Think about this for your own business.

How much of your stuff is so good that summarizing it with AI wouldn’t really save anyone, anytime, because there’s a lot to dig into, there’s a lot to understand, or is your content so thin that large language model could simply take it and extract the one bullet point of actual content, you have discard the rest.

And there’s no need for a machine, there’s no need for human to read your content because a machine can do it better and faster, and get to the tiny crumbs of useful information that are in there.

As marketers, we have to get better at creating valuable content.

As publishers, we absolutely need to create better content just to keep people’s attention just to hold on to the audience that we have.

So if you’re on the publishing side, and you’re publishing stuff that you know is not delivering and it frustrates people, now’s the time to reevaluate that, because your revenue model probably have to change really soon as machines become more and more adept at reading the web, extracting content from the web and presenting distilled versions of it to users.

That’s it for this show.

Thanks for tuning in.

We’ll talk to you next time.

If you’d like this video, go ahead and hit that subscribe button.

Mind Readings: The Real Danger to the News Industry Isn’t AI

Machine-Generated Transcript

Comments

Leave a Reply Cancel reply

Pin It on Pinterest