Human Centric AI - Behavioral Signals

Human-Centric AI: The Need for Empathy in These Difficult Times

Over the course of the last several months, we’ve learned a lot about how we work and interact with one another. Being forced to completely rethink the model upon which our economy is structured has given us a rare opportunity to evaluate potential improvements, changes, and optimizations that may have previously taken years to accomplish. 

While the circumstances around these transitions are tragic and there’s no guarantee that we won’t see further challenges ahead, many companies have accelerated their timeline for digital transformation to not only better respond to the current situation, but to be ready for anything that might come in the future. 

A big part of that process is the implementation of AI. While AI was a major focal point for most businesses before the pandemic, it now forms the foundation of digitization and optimization efforts in most major companies. But AI alone isn’t enough – it needs to be prepared for the kinds of challenges we currently face. Empathy is key and only a human-centric AI that augments, supports, and supplements human interaction can truly do that. 

Building a Link Between AI and Human Operators

AI is having an outsized impact on the shape of business. It already was before 2020, and its role in shaping our “new normal” has only accelerated that timeline. But, despite the long-mentioned concerns of AI replacing human operators or developing a means by which to provide human-like interaction with customers, businesses will still live and die by the performance and quality of their human resources.

AI cannot replace the instant empathy a human customer service agent feels when someone breaks into tears on the phone or the friendly. In fact, a significant number of researchers when asked by the Pew Research Center about the future of AI were pessimistic about the technology without some form of empathy. Nearly 37% noted the need for empathy in AI and several prominent leaders said that AI dependence requires careful attention paid to the way in which these tools are built, deployed, and maintained. In short, an agnostic approach without the benefit of a human touch could lead to more harm than good.

Building an ethical AI is a difficult challenge – one that many organizations are attempting to tackle. OpenAI is working on an ethical framework and Google’s AI for Social Good initiative is focusing on solving major humanitarian problems with AI. But at the root of success, many agree that we need what are essentially “soft skills” for artificial systems. A Natural User Interface (NUI) is built around a built-in intuition about human actions, both perceived and projected. 

We’re already seeing something approaching a NUI in the voice interfaces that are now present in every smartphone, digital assistance, and most desktop and laptop computers. These systems are designed to learn patterns in human behavior and adjust how they respond, and some are integrating emotion AI to evaluate and respond to subtle cues in the human voice that indicate certain emotions.

Going Beyond Trust in Artificial Systems

According to a recent Pega study on AI and Empathy, 40% agree that AI has an innate potential to improve how we interact with companies, but more than two-thirds of respondents said they trust humans more and would be more likely to tell the truth to a human agent than a machine. Part of the distrust is the long-standing fear that AI will take human jobs. 35% cited this concern, and another 54% highlighted what they see as innate bias in AI systems – driven by their programming. These are both significant, ongoing problems that developers have been working against for many years. 

For trust to develop with AI systems, we need to go beyond simple personalization. Knowing what you need when you need it is not empathy – it’s data analysis, and done improperly, it can be more unnerving than it is helpful. For AI to become truly empathetic, we need to approach it from a different perspective-one that takes into account Human Interaction (HI) and how it influences the actions and reactions of a machine. 

Building AI with the Human Experience in Mind

AI is not a perfect, sterile vessel into which we can pour data. It is created by humans and therefore carries many of the same biases and perspectives that humans do. Trying to create something that is “unbiased” is inherently flawed.

The development process needs to bring empathy as a key deciding factor for how the system will be developed and deployed. How will AI be implemented, who will it impact, and what interactions will it have with human operators? These are all questions that should be a fundamental part of the development process. 

Another major component is emotional intelligence AI – the ability of a machine to recognize key elements of human emotion and adapt its response accordingly. AI systems are being developed that can detect depression in a human operator’s voice, respond to raised stress levels, adjust responses based on elevations that occur during a conversation – such as anger on a customer service call – and more. 

By measuring human emotion throughout the customer experience, it is possible to go far beyond simply personalizing a greeting or recommending a purchase based on past behavior. Systems are able to build real relationships with individuals through carefully calibrated, highly precise responses to customers. This leads to improved customer service performance, higher customer retention rates, higher levels of customer loyalty, increased ROI across business units, and a legitimate competitive advantage over companies not utilizing this technology. 

The Role of Empathy in the Development of AI

Empathy plays a multifaceted role in the development of AI. From the teams assembled to design and build these systems to the way in which an AI system interacts with customers and human operators – empathy will be the game-changer that allows AI to have a lasting, positive impact on society. 

More than just a tool to analyze data faster, AI can become a constant companion to improve business productivity, improve customer experiences, and develop more sincere lasting connections.