Let’s be honest, on a day-to-day basis, does anyone interface with artificial intelligence (A.I.) cool enough to merit all the hype? Unless you own a self-driving car or work for the NSA, the answer is probably “no”.

But that is not to say that A.I. isn’t everywhere, because it is. It is at work whenever you get targeted with an ad online, when your phone autocorrects spelling mistakes, and when Facebook organizes your newsfeed. These applications of A.I. are helpful, but do not seem to merit the non-stop hype and fear mongering about robots taking over the world (or at least taking our jobs).

 
That is because very few people really understand what goes on behind A.I. and how to make it truly impactful. Behind A.I. are an array of other technologies and systems that have a far greater impact on the result.

“Artificial intelligence is only as smart as the data it is receiving, which means that the biggest gains you can make in quality will come from improving data input,” explains Vijay Chittoor, a thought leader in marketing A.I. and the founder and CEO of BlueShift. “A.I. in marketing technology is the same as A.I. anywhere else. Self driving cars are finally becoming a reality because of a massive amount of map and street view data that helps the A.I. anticipate every possible scenario. Professionals in any industry must prioritize higher quality data input to get A.I. out of the backseat.”

Here are three reasons A.I. is stuck in the backseat (and how to get it out):

Wrong emphasis

There are a gluttony of products available that claim to leverage the power of this new technology to do things never before possible: identify the food on your plate, match you with a romantic partner, or find deals on clothes.

Some of these funny tricks are entertaining (if unreliable), but all of them fall short of groundbreaking because their focus is wrong. Too many companies create their A.I. by finding a cool application for it instead of looking for a wealth of data that they can capture, organize, and generate authentic intelligence with. It is important to look at it this way, because powerful data sets will create powerful A.I. applications.

“Behind A.I. is some kind of machine learning, which has an incredible impact on the actual value output of the program,” explains Chittoor. “This is especially true in marketing. Today, the cutting edge of marketing technology is audience-of-one targeting, where you launch marketing campaigns using A.I. for individual consumers. In that case, A.I. is pairing a consumer with a campaign, but the real work was done by machine learning, which studied millions of events to fully understand consumer behavior. The better your data and machine learning, the better the A.I.”

Think bigger

If automation is really only as good as the data and machine learning behind it, why so much emphasis on it? The emphasis is understandable because when properly executed, A.I. should create groundbreaking outcomes.

But let’s set some ground rules for what qualifies as groundbreaking. A groundbreaking automation should remove the need for people to perform tedious or repetitive functions that do not uniquely add value. In other words, if it is a task that requires a person to mind-numbingly make basic decisions over and over again, A.I. should make a splash there.

“When you put A.I. in the driver seat in marketing, you are able to remove marketers from the role of button pushing,” Chittoor explains. “Launching campaigns and changing settings – that is glorified button pushing. Marketers are supposed to be creatives who develop powerful messaging concepts that win over consumers. But today, many marketers spend huge volumes of time on tasks that can and should be automated.”

Goldrush mentality

In the rush to release an A.I.-something, many companies have simply developed gimmicks. Silicon Valley’s “Hot Dog App” that helps you discern whether or not the food on your plate is really a hot dog is satirical gold for that reason. For the last couple years, all anyone had to do was breath the word A.I. to an investor to get funded.

Out of all of that, focus on quality and the real value of A.I. was missing. Only a select handful of companies – in the B2B or B2C space – have A.I. platforms that merit the hype. Here are a few tips to identify them:

Do they talk about data? An A.I. company should be selling its theories on data capture and analysis, not just its automation.
Is the outcome groundbreaking? This means you should be looking for solutions that remove tedium and time waste, not just cool tricks and flashy branding.
Is it a front seat application? Today, A.I. can do important tasks. Programs that are limited to “helpful alerts” and push notifications are way behind.