You Ask, I Answer: Impact of AI on Content Marketing?

In today’s episode, Brian asks, “What sort of impact have artificial intelligence and machine learning have had on content marketing?

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.

Christopher Penn 0:13

In today’s episode, Brian asks, what sort of impact have artificial intelligence and machine learning have had on content marketing? Gosh, there’s so many answers to this question.

But they, they boiled down to three things.

Regression classification generation.

So first part regression analysis, AI and machine learning have done an incredible job of taking data that we collect about our content and telling us what works, right, being able to say, okay, these channels, are these tactics, or these even pieces of content have worked.

And there are a legion of mathematical techniques that answer those questions that that can demonstrate what really works and what doesn’t and savvy marketers with good in house data teams or with the right combination of vendors and agencies can use that information to to in their efforts to say we’re going to do more of X and less of why we’re going to create more of these kinds of videos and less of these kinds, and so on and so forth.

So that’s part one.

Part two is classification.

This is especially important for doing research for content marketing, we have more data to go through than ever before, right? We have more content of other people’s to try and understand even though I was in the midst of doing some topic modeling earlier, for a client, trying to figure out out of 90,000 conversations about this client on Reddit, what are the relevant topics so that we can create a content strategy around the things people are talking about? And using classification techniques to sort of bucket and sort and organize this information, because I can’t read 90,000 Reddit posts in a reasonable amount of time, remember everything in them, and then distill them down into something that I can create a strategy around, but a machine absolutely can do that.

And then from there, I can say, Okay, well now, based on this information, what? What’s my strategy, right, so that’s part two.

And then part three is generation.

And this is where content marketing is going to be in a very, very interesting place in the next few years.

Deep learning models like GPT, three, or GPT, Neo x, or any of these, these really large language models, the various models of the BERT family can create text, they can write, based on prompts they’re given, and parameters and guidelines.

And right now, they can generate mediocre content.

Like they can write press releases, like nobody’s business.

And they’re coherent, they’re sensible.

They make logical sense.

They’re readable, but their press releases, right.

These are not pure howitzers.

By any means.

And this is a big improvement from three years ago when machines wrote garbage.

Right? They wrote a nonsensical word salad.

In another three years, these machines can probably create good content, not great, not amazing.

Not Pulitzer Prize winning, but a step above mediocre.

And what’s happening in the startup field in content marketing is you’re seeing an enormous number of companies appearing that are offering these services, hey, we can generate a million blog posts for you, they will all be mediocre, and relatively uninteresting, but the hit the right combinations of words and phrases that are coherent, and average.

One of the challenges of artificial intelligence, machine learning and marketing is that the power law curve, right? There is relatively little great content.

In the big picture.

There’s a whole bunch of mediocre content.

And there’s a vast ocean of crap.

And so when it comes to training, machine learning software and hardware on what content to create, the pool, we have to choose from the content we’d like to create great content is very, very small.

And so a lot of these companies have trained their machines, their software, on that middle pool of mediocre content because there’s enough of it for these pieces of software to work and we are seeing with These newer models, the ability to fine tune them, to essentially use the great content to nudge them in a direction away from the pure mediocrity they were trained on.

Christopher Penn 5:13

But it’s going to take a while for a lot of our content generation algorithms to be able to work with the relative sparseness of great content.

And to be able to generate it on their own.

Think about it.

Think about your own blog, or your own newsletter, or your own podcast or YouTube channel, whatever, how much of your content that you’ve created, would you say is great, really terrific.

Love it is done well, big numbers.

How much of it is okay, I got off my to do list.

And how much it was, well, that was embarrassing.

Let’s make sure though, no one ever sees that.

It probably very little ladder.

Hopefully, it was caught in the editorial process and never saw the light of day.

But probably most of your content is in that middle bucket of I got it done, I got it off my to do list.

Not I made something great.

And because of that, our machines are going to struggle for a while to replicate great, because there’s just not enough examples of it.

Which means that for you, as a marketer, you still have an opportunity to have a long and rich and fulfilling career as a content marketer if you create great content consistently and frequently.

And if you can create great content consistently and frequently, you will remain ahead of the machines because there are not enough people like you who can make the content that these machines can learn from.

Right.

That’s the impact that machines have had thus far.

They can do regression like crazy.

They can do classification really well, and they can create mediocre content.

And they will eventually evolve in the next few years to creating good content.

But it will be a while before they create great content.

So that’s the impact


You might also enjoy:


Want to read more like this from Christopher Penn? Get updates here:

subscribe to my newsletter here


AI for Marketers Book
Get your copy of AI For Marketers

Analytics for Marketers Discussion Group
Join my Analytics for Marketers Slack Group!


Pin It on Pinterest

Shares
Share This