Imagine having a conversation with Alexa where she understands everything you say and feel—even when you’re being sarcastic, or when you sound frustrated or sad. Amazon is currently working with thousands of engineers and college students to help her do just that. Why? Because creating AI that understands not just words, but nuances of emotion will allow Alexa to act less like a friendly robot—and more like an actual friend.
Indeed, Amazon isn’t the only one with the goal of helping AI chatbots become more human. While many of us are still struggling to write voice texts without having to edit them—some of the smartest engineers in the world are working on the next wave of conversational AI capabilities, including reading and understanding human emotions. The advancement would take AI to a whole new level. Moving forward, Alexa won’t just take directions, like playing a certain song or setting your alarm clock. She’ll be able to anticipate when you need encouragement—recommend calling a friend who can help—or determine when a certain product or service might help you. Chatbots will be more than a fun novelty; they’ll become a meaningful part of our everyday lives.
As conversational AI continues to improve itself, it becomes obvious that emotional injections are the logical next step forward—and the sooner the better. Research shows chat tech like Cortana, Siri, and Google Now can be incredibly insensitive when it comes to user queries on ailments and personal issues. That’s because as machines, they lack the empathy and compassion needed to properly respond to those issues. In her recent article “AI and Chatbots: Getting Their Sensitivity Training,” my colleague Shelley Kramer discussed Koko, an MIT-born machine learning project that could help infuse any AI bot with the empathy users need to feel understood on a personal level. No doubt many other engineers are working hard to offer a similar service, allowing tons of new bots to adopt the characteristics of humankind easily.
If done well, the technology could be incredibly valuable on a personal—and business—basis. The following are just examples how.
On the Personal Side
If chatbots can be engineered to read and respond to the nuances of human emotion, a whole new wave of service opportunities could open up to them, and to those who need them. For instance, chatbots could perform talk therapy, making mental health services even more affordable and widely accessible throughout the country. Chatbots could be trained to understand the nuances of those who are paralyzed or speaking with variant speech patterns, giving them more freedom and the ability to better communicate with others on a daily basis. And in a hospital environment, bots armed with both emotional and multilingual skills would be especially useful in translating sensitive information to patients and family members. Imagine the benefit to patients and healthcare workers alike!
When it comes to personal assistants, emotionally intelligent AI would move the possibilities to a whole new level. They could help talk us down after an argument or prevent us from making bad decisions in the heat of the moment. Think C-3PO with a jetpack—the options are literally endless.
On the Business Side
Face it: companies wouldn’t be investing millions into emotionally intelligent AI if there were no business upside. By and large, those benefits will be found most immediately in improved customer service—but they definitely won’t end there. As I shared in my piece “Artificial Intelligence: To be Feared or Embraced?” last month, a study from Oxford University estimated some 50 percent of U.S. jobs will be obsolete in the next two decades due to automation and deep learning. Emotionally intelligent AI Is a huge part of that equation.
Why? Besides customer service chat bots, which are already making their way to the front lines of businesses everywhere, empathic bots can be trained to perform lead generation, research and recommend products, and even pitch some products directly to a user based on need. For instance, if a chat bot knows a user is unhappy with a certain product or experience, they can refer them to another company or recommend a product they might like better. They can make product recommendations based on where someone is traveling and what the weather may be. And even better, they will have an entire world of IoT product data and information in their “brains”—far more than any human customer service agent could hold—to make those recommendations.
If you’re like me, you might be wondering: OK, so who is my bot loyal to—the business landscape or me? That’s one of the greatest questions that has yet to be seen, in my opinion. Just as they are able to serve humans, emotionally intelligent AI also have the ability to capture and process lots of personal information and transfer it to companies instantly. How will this data be shared? Who will profit from it? Those are issues that still need to be addressed as the ethics side of AI seeks to catch up with the technology itself.
Additional Resources on This Topic:
This article was first published on Futurum Research.