AI Is More Creative Than You Think


July 22, 2019 - 0 Comments

Artificial intelligence does three things really well: acceleration, accuracy, and automation. Today’s machines are good at accomplishing tasks within these domains, but they aren’t yet capable of crossing them. They stay in their lane — the lane their human told them to stay in. As a result, we view machines as being better suited for “left-brain” tasks. But more and more, artificial intelligence is bridging into realms once considered uniquely human because they are right-brained.

According to my friend Christopher Penn, co-founder of TrustInsights and author of AI for Marketers, there will be two types of marketing jobs in the future: Either you will manage the machines, or the machines will manage you. This includes creative tasks, which were once considered a no-go zone for AI.

In creative disciplines, much of the work is procedural. If you’ve ever set up a social media ad campaign, you know this. If you’ve ever written headlines or email subject lines, you know this. To the extent creative work can be templated, it’s up for AI grabs.

In 2016, 20thCentury Fox partnered with IBM Research to explore the idea of whether or not artificial intelligence (Watson, in this case) could create a movie trailer. The movie was — not ironically — a horror / suspense film about artificial intelligence. IBM taught Watson how to understand the relationships between the patterns and types of emotions humans consider frightening, suspenseful, and scary.

The result isn’t bad.

Sure, it’s a little amateur. But it looks like something a human amateur could create. And you know how humans work: we get better with time and practice. So does AI.

When it comes to AI and machine learning, right-brain tasks are absolutely on the table for the future of marketing, and I believe it can be powerful and impactful. Here are four ways AI can free up your creative resources to do more strategic work not in reach of the machines.

Testing Social Ad Creative

Today, the creative tasks most within reach of AI are those that are templated, procedural, or even formulaic.

Like social media ads.

Sure, there’s human magic to selecting the ‘right’ creative and pairing it with the ‘right’ message, headline, and copy, or even inventing the idea of an ad campaign to begin. But we make creative decisions based on our own biases and assumptions and usually end up running  A/B tests to determine which of our assumptions get the greatest response. We learn from those tests, and then we apply what we learn to the next campaign or optimize the one we’re already running.

How many variables can you test for? Three? Six? Maybe twelve?

Pattern89 is an AI platform for paid social that helps marketers discover which of the 2,500 different dimensions about their paid social programs drive performance.

Yes, 2,500. That’s a lot more than 12.

According to R.J. Talyor, CEO of Pattern89, most marketers want to test, but by the time they get around to it, they don’t have the time to create and launch the test.

You’ve probably heard that data is the lifeblood of artificial intelligence. Without it, machines are unable to get better because they can’t improve. At Pattern89, your ad creative is tested against Pattern89’s algorithm, which is powered by machine learning and has been learning from a pool of ads over time. When you use the service, you’re testing your ad creative against a database of ad creative. The result? Specific feedback about precisely how to optimize your ads. It’s almost like getting an ad creative recipe. Your role in all of this is to be the human and make something great from that recipe.

Using Emotion to Tailor Messages

Artificial intelligence is all about getting a machine to replicate human intelligence features, and there’s nothing more human than assessing someone’s emotional state and rephrasing our messages or ideas to suit. Certainly, this can’t be done better by a machine, right?

Apparently, not.

Persado is a company I’ve been monitoring closely since doing an early demo with them more than five years ago. They use a dizzying array of data inputs and DMP relationships to create highly targeted, massively specific content pieces inside email, on a home page or landing page, and for social media. Based on users’ historical engagement with messages of different tones and tenors, Persado can combine copy and images and Frankenstein together the perfect ad for EACH USER.

The language for these messages is characterized according to 15 different emotions such as ‘gratitude, ‘achievement’ or ‘fascination.’ Here’s their complete list:

 

Image Credit: Martechtoday.com

 

Each message is created on the fly or in real-time, which simply boggles the mind.

Anyone who has developed an influencer marketing program knows the work is part art, part science. Sure, there’s a method to the madness when evaluating and selecting influencers. But our bias or lack of time gets in the way of forming influencer programs that take into account the myriad of ways we can actually reach our audiences through influencer programs that are more creative or innovative than standard. When it comes to influencer programs, defaulting to the influencers perceived to have the greatest influence is faulty. TrustInsights recent 2019 Social Media Influencer Benchmark study found that for Facebook, Twitter and Instagram, smaller influencers get more engagement than the big-time influencers unless The Rock is part of your influencer program.

The finesse and fine points of influencer marketing are further exacerbated in the fact that influence is not created equally across all social media platforms, making it more and more important to comb through your influencer options, but who has the time?

Again, enter AI.

AI is transforming influencer marketing. Platforms like Linquia and Influential are already using it to help marketers understand and select influence, weed out fake engagement, and predict what incentives to use for individual influencers.

Image Credit: Single Grain

Machines aren’t taking your creative jobs, but they are creating a giant opportunity to augment human creativity.

R.J. Talyor, CEO of Pattern89 described it best when he said, “I think photographers, illustrators, artists and marketers who are artists in some way are required to make brands feel and talk and seem different. Machines, if left to their own devices, will make us all sound the same.”

 

Jay Baer is the founder of Convince & Convert, a Hall of Fame keynote speaker and emcee, host of the award-winning Social Pros podcast, and the author of six books including Talk Triggers: The Complete Guide to Creating Customers with Word of Mouth.

Read more of Jay’s Blogs here!



Leave a comment

We'd love to hear from you! Your comment(s) will appear instantly on the live site. Spam, promotional and derogatory comments will be removed and HTML formatting will not appear.