How To Make Use Of AI: Why Marketers, Consumers And Workers Should Trust Artificial Intelligence (Opinion)

Few of us look at artificial intelligence as a new, brilliant coworker. Instead, we see AI as the cyborg in the first “Terminator” movie: a futuristic menace who's come to wreak havoc — and take our jobs.

But if companies want to boost productivity and better serve their customers, employees must get educated about AI and embrace it.

AI is a subset of “machine learning,” the broad term used to describe machines that learn to perform certain actions based on data sets. For instance, a self-driving car operates via machine learning. It follows a set of rules that define how it should drive: stop at a red light, signal on a turn, slow down at a school district, etc.

AI occurs when a computer mimics human cognitive behavior — such as understanding language and rewriting content. Siri is an example of AI. A perhaps less well-known example is the AI program the Associated Press uses to cover minor league baseball games and write earnings reports.

The newest and hottest version of AI, which is better known as “deep learning,” occurs when a program not only understands language and images, but can interpret them and respond to them on its own. Google Translate, for instance, now uses deep learning to apply irony and slang to its translations. And one health care startup, Infervision, is creating a deep learning program that can examine x-rays and identify early warning signs of cancer.

With this range of skill, it’s no wonder people are scared AI will take their jobs. After all, haven’t we always thought of writing, understanding slang and interpreting pictures as uniquely human skills? Indeed, 70 percent of consumers across the globe are fearful of AI. Meanwhile, 34 percent of workers in a variety of fields say they are “afraid of the concept of AI.” As a result, only 17 percent say they are “actively using” AI, while 49 percent say they “have no plans for any use of AI.”

This reluctance to aggressively adopt new technologies is one reason that “growth in business productivity ... stands at its lowest rate since the early 1970s (1.3 percent),” according to a new Deloitte study.

But AI and deep learning can greatly assist us in our work. It can free financial analysts from writing run-of-the-mill earnings reports so they have more time to probe a company’s books or investigate new risks and opportunities facing a firm. It can help radiologists catch cancer earlier, so they can spend more time counseling patients and planning a treatment regimen.

AI can also help marketers better target consumers and alert them to the goods they desire.

Consumers and retailers alike are familiar with the old “spray and pray” marketing strategy and the damage it can do to a company’s brand and growth. Blasting a demographic with generic ads, countless emails and pop-ups can backfire. Customers, feeling inundated with impersonal messages, simply unsubscribe from the company’s email list or block all ads.

The use of artificial intelligence can spare consumers from this daily barrage of poorly aimed marketing, allowing personalized campaigns targeted to specific consumers.

But marketers have to use it correctly. Right now, many companies that use AI try to control AI’s powerful algorithms by suffocating them with rules. That defeats the whole purpose — deep learning has so much potential precisely because the programs are able to self-adjust based on real-world data in real time.

These restrictions cause machines to falter and marketers end up making the same mistakes — sending emails or ads for products that consumers already have bought or have no interest in. A recent survey found that nearly half of all consumers receive emails that are “rarely” or “not at all” relevant to their current wants and needs. That leads to dissatisfaction and plummeting sales.

Unleashing the power of deep learning could help marketers personalize ads in ways they never could before. With fewer rules and more intelligence, a computer can discover connections, correlations and causations that a human might never notice.

For instance, the United Services Automobile Association recently started using AI to predict how customers might get in touch with the company and for what reasons. The machine makes these predictions by analyzing thousands of data points and applying broad customer patterns to individual members.

With the help of AI, USAA improved its “guess rate” from 50 percent to 88 percent. Now, USAA customer service representatives can better predict when a customer will contact them and for what reason. That knowledge allows workers to better prepare to serve the customer.

Humans need to recognize that machines are on their side. They’re not trying to steal their jobs — and they still answer to us. Remember that cyborg in the first Terminator? He came back in the sequel not to destroy mankind, but to save it.

Lars Fløe Nielsen is Chief Development Officer of Sitecore.