As part of the AI Academy for Marketers membership, we offer monthly Ask Me Anything (AMA) coaching sessions with leading industry experts. We chat about everything from technology trends, use cases, lessons learned, to much more.
This month's AMA features R.J. Talyor, CEO, Pattern89. During the conversation, R.J. discusses Pattern89 and its services, AI in advertising, machine vs human roles, and the ethical risks of AI—all stemming from his course in AI Academy (linked below).
Below is a quick video from our chat, followed by top takeaways from the conversation.
How does Pattern89 learn and know what specific creative dimensions will perform best?
More than 6,200 brands are connected to Pattern89. When we connect any new customer, we ingest the creatives they've used and the performance to date. It looks at the full picture and applies computer vision and Natural Language Processing on top of that to understand what’s in that image. From there, it digs into the click-through rate, conversions, and cost per impression. Pattern89 does that for every creative and for every customer. Every 24 hours, that data analysis is re-done to understand what is trending up (and down) within creative dimensions. For example, the machine can identify that the color red is causing ads that feature couches to be more expensive (the algorithm is de-prioritizing the color red in that instance). The data spots trends in the creative in terms of what is trending up and what works (across emojis, nouns, verbs, etc.). We use that data to make predictions that are over 95% accurate!
Stemming from that question, if you are pulling off past creative and its performance, how does the AI layer in new trends in design/art to make recommendations?
Machines cannot come up with new ideas. That is a human's job to come up with wild, wacky, outlier ideas and find what's new and creative. The machine’s job is to then say yes, that’s a good idea or a bad idea from a metric’s perspective. I want to be clear though that the machine is not always right. It's way better than A/B testing. The machine will validate ideas and predict performance but it won't create the next new global-changing campaign. Humans will come up with those ideas. Humans will be held accountable for creativity and strategy.
Will we ever get to the point where AI machines will be able to create ads?
Yes. But is it art? The machine is not acting off emotion or shared experience or memory, operating off an algorithm. Machines cannot understand the nuance of the creative choices they make. Machines don’t know the context of the political environment, for example. We have to apply the human filter. It has to be someone who understands the larger context. There is so much brand risk for the wrong thing that slips through.
What are the ethical risks?
Tons of them. The problem is too big for anyone to get their head around. I'd suggest three things.
Since AI tech like Pattern89 predicts success based on previous success, wouldn’t most creative end up looking the same or similar to other brands using AI tech? How do brands stand out, while also aligning with optimized predictions?
The machine honors your data first. So what does your creative look like? It will identify outliers to try. This is where there is a component of a recipe. In the food world, you can give a recipe to three different chefs and get three different meals. It’s the marketer's responsibility to create content from the recipe they get from Pattern89 and determine how it fits into the customer journey. There is the opportunity for the human to come in and make those changes. Not everybody is getting the same predictions bc the base models are not the same.
Become an AI Academy Member Today
Learn more about the topics discussed in this post with a membership to AI Academy for Marketers. The Academy is our members-only online education platform and community that features dozens of on-demand courses and certifications taught by leading AI and marketing experts.