What if the moment you go into a store or log on to your favourite e-commerce site, you were greeted by a human-like assistant who’s there to answer your inquiries with a warm and expressive smile? This interaction would be different from that of today’s chatbots or Amazon’s cashier-less Go stores that have no need to engage human employees and rely entirely on typed text interactions.
Today, we’re on the cusp of the arrival of digital human assistants that will fill our need for helpful and knowledgeable interaction expressed with those oh-so-human qualities of voice, emotion and realistic facial expressions. Australian-based company Soul Machines is working on 3D human-like faces that learn via artificial intelligence (AI). But what these expressive digital human assistants are learning is not just what colour you may prefer for your dress shirt, but also what expressions and voice inflections go along with key conversation inputs to facilitate that very human desire to make a connection.
Consumers still want human interaction
Technology has changed how we shop. Previously, a retail representative would help customers navigate through an array of choices and help process the eventual transaction. But thanks to e-commerce and point-and-click shopping, there’s no need to engage a real person to buy that new furniture piece or order that comfy pair of socks. But with all the scrolling and clicking, many have wondered if we’re on the right path. There’s even evidence that consumers want more human interaction, and it is often that very same interaction which solidifies that holy grail of marketing — providing a unique customer experience.
Seventy-five per cent of consumers around the world desire more human interaction (according to PwC data from a total of 15,000 global respondents). This survey also points to an increasing level of fickleness, as one in three customers said they would walk away from their preferred brand after a single bad experience. With so much riding on a company’s ability to provide a positive customer experience, it is no wonder that companies such as Autodesk, Conversagent and IBM are trying hard to harness the power of AI, machine learning and natural language to deliver that humanlike interaction — minus the human.
What’s making it all possible?
It admittedly sounds like science fiction, but the important pieces are already in place. Big data, petabyte-scale processing power and machine learning are all making it possible to create digital human assistants today. And because the backbone of this collection of technology is based upon AI, these digital human assistants will only get better as they learn.
For instance, natural language understanding and sentiment analysis are vital pieces in creating digital human assistants. Understanding what the customer is saying is just one piece. Knowing how the customer feels during the conversation is the new twist. With these two capabilities, the digital human assistant can personalize its replies and appropriately respond to the customer throughout the entire experience. But for natural language understanding and sentiment analysis to work in this manner, enough data needs to be available so the computer powering the digital human assistant can both derive patterns and learn. And because all of this interaction needs to be in real time, massive amounts of processing power have to be utilized so that the digital human assistant never slips up with painful and unacceptable lags or delays.
What makes digital human assistants such a massive leap ahead of today’s chatbots is that they don’t rely on text-based interaction, but instead provide a visualized human that the customer can both see and interact with. Voice and speech can be created specifically according to language. Accent requirements and movement of the facial muscles, eyes and mouth are made to match both voice and emotion to make the entire experience seem as if two people were conversing.
What does this mean for businesses?
Digital human assistants could have a huge impact across different industries. Banking and insurance often deal with complicated policies and regulatory procedures. With a digital human assistant who is always available, never sleeps and can cheerily help a customer navigate the complexities of these industry products and services, much of the friction of these types of transactions could be reduced or eliminated.
In healthcare, digital humans can provide 24×7 access to cognitive behaviour therapy that helps people monitor their moods and learn about themselves. The digital human assistant’s capacity to learn, express and emote could be a positive feedback mechanism for patients who benefit from this therapy.
We’re certain to see advancements made in digital human assistants — most likely sooner than we think. Top neuroscientists, AI researchers, psychologists and companies are busy working to bridge the gap between the capabilities of computers and the human capacities to learn and express. The impacts could be far-reaching, not only for business but for society as well.