Can AI help digital experiences feel more human?

Whenever we hear the words “artificial intelligence” it’s hard not to imagine a dystopian future where robots regularly get mistaken for humans, and we never have to wash our dishes, drive a car, or mow the lawn. It’s easy to get swept away with the futuristic possibilities of a machine that thinks and learns for itself—especially one that does your laundry. But given that we’ve yet to hear about any robots doing household chores (with the exception of the Roomba) what’s all the AI fuss about? Better yet, how does it impact everyday consumers and the people who design those products and experiences?

User experience, while ultimately grounded in the idea that a real human stands behind a brand or product, no longer requires an actual human to deliver that experience. If you're not in the mood to chat with a friend, you can text them instead of calling. And if you're not feeling like dealing with the crowds to go out to eat? You can order dinner with an app on your phone. And no one ever has to be embarrassed about getting lost again. Google Maps can save the day, no judgments.

All those interactions are powered by some level of AI. And while it may seem counterintuitive, AI might be one of the most important tools in a UX designer’s toolbox when it comes to humanizing the customer experience. In fact, industry analysts like Gartner predict that AI will drive over 85% of customer interactions by 2020. Let’s take a look at a few examples of how AI is being used today, and how it helps facilitate better experiences.

Chatbots

Chatbots, sometimes referred to as conversational AI, strive to mimic real human interaction via digital interfaces like an app, website, or kiosk. Have a question about a charge on your credit card after hours? A chatbot on the bank’s website can help you out. Feeling under the weather and not sure if you should go to the doctor?

A healthcare chatbot can help evaluate your symptoms and make a recommendation for treatment. Advancements in AI have improved these interactions to the point that some customers can’t tell that they’re interacting with a machine and not a person—that’s an important achievement considering the threat of the widening empathy gap with digital experiences.  

Call centers

We all know the pain of getting stuck in an infinite loop with a recorded message when trying to call our bank or internet provider for assistance. There are only so many times one can tolerate hearing, “Press one for your account balance, press two for….” with no real option to speak with a human being. You’d think that would make call centers an unlikely candidate for AI, right? After all, isn’t it the human interaction we’re missing in this experience? There’s one thing a computer can offer us that a living, breathing person can’t: consistency.

No matter how much training call center employees are given, they’re all unique individuals. How one employee handles a frustrated customer will be totally different than another. There’s also the human factor to consider. Those employees are real people, with real emotions and personal experiences that will guide their attitudes and actions.

Every employee will naturally provide a different experience for each customer—good or bad. An automated system with AI, however, can access information and provide answers with consistency that would be impossible to provide from a group of unique individuals, not to mention do it faster and with higher volume, in as many languages needed. Improvements to voice recognition and natural language processing are also helping these systems better understand callers and get to the root of what they really need.

Personal assistants

There are some things we all need help with, but don’t necessarily have the means or time to employ a real person to do them. Consider what it would cost, for example, to employ an assistant to schedule all your meetings. You’d have to train them to learn your schedule, preferences for meeting length and time. They’d need to know how much time to give you between meetings if you’ll be at opposite ends of the city during rush hour, and how to prioritize meetings between individuals.

A real person can do this, but so can Amy. Amy, the creation of x.ai, is a digital personal assistant that utilizes AI to learn your schedule and preferences so she can handle all that tedious meeting scheduling you used to have to do yourself. She sends out the invites, interprets responses and automatically suggests alternates that meet everyone’s needs. Amy’s sophisticated machine learning algorithms enable her to react to everyone interacting with her in a surprisingly human way. I sent a test email to schedule a meeting but purposely didn’t specify a location and this was Amy’s response:

But Amy can’t do it all on her own. There’s a real “teacher” (a team of AI trainers, actually) behind the scenes that help Amy with her machine learning curriculum. Amy can capture and interpret more data and at a faster pace than a person, but she needs the input from a real person to add that human touch to her repertoire. UX would naturally be a part of this process, especially teaching Amy to recognize signs of frustration and identify opportunities for empathy and compromise. Amy is trained to take cues from her programming, her trainers, and the person she’s assisting, to respond appropriately to requests. If she wasn’t given any human feedback, she’d risk running into a “Sorcerer’s Apprentice” situation. x.ai’s CEO and founder, Dennis Mortensen explains,

"Suppose we tell a self-driving car to get us to the airport as quickly as possible. Would the autonomous driving system put the pedal to the metal and drive at 300 mph while running over pedestrians? Very much in the same vein, if I tell Amy to 'set up a meeting with Matt as soon as possible,' would she push a reminder every 10 minutes until Matt responds? No, and that would obviously not be the outcome you had in mind when asking her to do the job."

SVOD recommendations

Whether you’re a movie buff or watch a flick only occasionally, if it wasn’t for streaming services like Netflix, HBO, Hulu, and Amazon, you’d likely miss out on hundreds of titles you’d never consider without their recommendations. Every time you search, view, watch (and probably a million other data points), the AI in the background for these services is continually learning about your preferences, and even predicting your behaviors, creating a smooth, enjoyable experience.

Yet once again, these experiences are a joint effort between both AI and UX. UX Magazine notes that it’s not the AI—or rather, what is recommended—that endears Netflix to users, but how those recommendations are presented. You don’t just get recommendations for comedies, you get a recommendation for dark comedies. That level of personalization humanizes Netflix’s AI, helping blur the distinction between man and machine.

More human than a human?

So is AI the future of UX? While the concept and results are impressive (and just fun to think about) AI can’t exist without us humans. Right now, in its infancy, AI is a tool we can use to better our connection with users. We then can design better, more efficient, and adaptive experiences that engage with users and develop trust and loyalty. A robot revolution is (hopefully) a long, long way away.

But the opportunity to collaborate with AI is happening now. When machines can pass themselves off as human, we can start worrying about that robot revolution. Until then, AI that’s designed with empathy and lots of user testing will be more helpful, effective, and will look nothing like the apocalypse. AI is already here. But without a great experience with these interfaces, it could easily lose its spot as an icon of the future and become a thing of the past.