Issac Asimov predicted that robots would cause problems. Not that they would try to take over the world, but that they would take over jobs. People would be unhappy about this.

We’ve already seen this happen with the advent of assembly lines at factories where robotic arms can assemble whole cars. But these are dumb machines.

What happens when a machine can create quality marketing campaigns? What happens when the machine gets better than the human at keyword research? Will the digital marketing agency become a thing of the past?

Fredrick Vallaeys, former Google employee and co-founder of Optmyzer, doesn’t think so. Here’s why.

There Are Things Machines Don’t Know

Machine learning already seems to be able to learn about the world on its own…sorta. It can mimic human speech…sorta. And it can predict the future…sorta.

But what it doesn’t yet have is intuition, a distinctly human trait. What is intuition? It’s the ability to take seemingly unrelated pieces of information and stringing them together by their common threads.

A good example is one that Fredrick himself brings up. His team decided to look at lunar cycles to see if they influenced predicted click-through rates. (It didn’t, btw.) And a machine today wouldn’t be able to make that sort of abstract connection between human behavior and lunar gravity.

Machines Don’t Have Nuance

Human language is subjective most of the time. We rely on context clues to communicate. Machines can’t always pick up on these things and they won’t be able to service clients directly until they do.

For example, a veterinary clinic might have a sudden influx of clients during the summer. If they don’t look at the actual numbers, they might just tell you they’re full-up and don’t need more. But what does that mean? A machine would see that as 100% capacity, no need for marketing.

A good marketer might ask, “well, really how full are you? What percentage increase is that and what percentage of available time are you filling? We might be able to bring on more clients if we market better. Then you might be able to afford more techs and your capacity could increase.”

A machine couldn’t advice a client on these issues. All it can do it take human communication literally.

A Machine Can’t Empathize

Nobody wants a robot when it comes to bedside manner. We want someone who understands out pain when we’re sick.

Similarly, when a campaign goes south, we want to know why. We want someone to explain how it’s going to get better or how things are going to get fixed. At least for the time being, a machine can’t do these things. It can only run a program and spit out data.

The Machines Owners Are Rich (and Will Be For Sometime)

Whoever trains the robots will control how they give advice. Google is a big player in machine learning. Their search engine is one of the largest set of ML programs on the planet. And they’re rich.

To them $1000 is nothing. To a small business it’s a day’s revenue sometimes.

If you have rich, disconnected people teaching the machine, you’re going to get a machine with a disconnected mindset. And when it comes to setting campaign budgets and making recommendations based on those, you want to avoid a disconnect between client and service.

When people who understand SMBs offer services to SMBs, they’ll outstrip ML solely based on empathy toward budget. This is why AI won’t take marketing jobs just yet. They still can’t add the human element.

By Ben Mattice

Benjamin Mattice is a freelance writer/editor, horror and sci-fi writer, SEO and affiliate marketing newbie, dog wrestler, cat wrangler, capoeirista, and long distance runner. He lives in the Palouse with his wife, three dogs, two cats, and two rats. Yes, that would probably be considered a mini-zoo.