Will Your Next Copywriter Be an AI Program?

In July of 2020, JPMorgan Chase & Co announced that it will now assign copywriting tasks to artificial intelligence programs. The financial services firm will use the technology for email marketing. JPMorgan Chase & Co plans to implement tools from the software company Persado within the next five years. Marketing executives are under growing pressure to demonstrate and improve the results of their work. Some hope that AI can help on both fronts, even if it means playing a smaller role in certain creative decisions. Let’s take a look at some AI copywriting applications to see how well they perform. 

AI vs. Human Copywriters 

Since we mentioned the Persado program, let’s look at it first. According to the Wall Street Journal, JP Morgan Chase asked human copywriters and AI to write headlines for a promotion they were doing. Humans came up with “Access cash from the equity in your home,” with the call to action “Take a look.” A variant created by Persado was headlined “It’s true—You can unlock cash from the equity in your home” and suggested “Click to apply.” The version created by the AI system generated 47 weekly applications for home equity lines of credit, compared with 25 for the human-created version.

While these are impressive results, this is only one test. Still, when we look at a program like GPT-3 which offers API integration and can return a completed text, attempting to match the pattern you gave it. You can “program” it by showing it just a few examples of what you’d like it to do; its success generally varies depending on how complex the task is.This program has more than 175 billion parameters and it is worth mentioning that GPT-4 will come out soon which will have 100 trillion parameters, which is 500 times the size of GPT-3. With such a rapid pace of development, we can expect to see machines competing with humans in terms of content writing in the near future. 

How are Such AI Text Models Created? 

Programs like GPT-3 and other AI text models rely on training data that was prepared by human data annotators. To get a better understanding of text annotation, let’s take a look at the image below: 

Image Source

In the text above there are proper nouns like the names of people and countries. There are also more complex grammatical structures such as words linking anaphors and cataphors to their antecedent or postcedent subjects. All of these things would need to be annotated. For example, annotating proper nouns falls into the category of named entity recognition. Linking anaphors and cataphors would be more of discourse annotation or part of speech tagging. There are many different types of text annotation that need to be used when creating the types of text systems we talked about earlier. These massive amounts of data need to be annotated with extreme precision. 

Will AI Systems Replace Human Copywriters? 

In the near future, AI and human copywriters will work hand in hand. For example, AI writing systems can speed up the writing process and help with repetitive and tedious writing tasks such as creating slight variations of the same ad generating proposal and email templates, and many other things. 

It’s also important to understand where the disruptions will actually occur. When a new technology or movement starts taking hold, it doesn’t do so across the board. It goes after the low-hanging fruit first, but then it takes a long time to move beyond that level.

The real impact of AI in the copywriting world will be in organizations like Chase Bank — businesses that produce and manage massive volumes of marketing assets. We’re talking tens of thousands of pieces of marketing collateral across multiple platforms. When you’re dealing with that kind of volume, having technology that can reliably analyze complex performance patterns and crank out new copy fast makes sense.

Huge, established brands are also much better candidates for copywriting AI technology than other businesses. In other words, companies where the brand itself is the main driver of results, as opposed to the copy.

Trust Mindy Support With All of Your Data Annotation Needs

Whether you’re building the next GPT-3 or any other AI system that requires data annotation, Mindy Support can assemble a team for you that will take this work off your shoulders. We are the largest data annotation company in Eastern Europe with more than 2,000 employees in eight locations all over Ukraine and in other geographies globally.  Contact us today to learn more about how we can help you.