AI visionaries: Xiaoming Liu on striking the right human/AI balance

Customer support experts are crucial to creating and sustaining an emotional connection between your customers and brand. They must display emotional intelligence (EI), an essential ingredient in creating authentic brand intimacy, and a uniquely human thing. Emotionally intelligent engagement makes customers feel seen and ultimately drives business results because customer support experts can go beyond the surface and understand and address what customers really want.

But experts can’t do it without help. The proliferation of omnichannel and ever-changing customer demands means that customer support experts have more responsibilities than ever. It’s up to the organization to provide the right tools — such as AI — to deliver exceptional customer experiences.

Yet even as AI deployment can bring about significant customer experience improvements, concerns persist, on both the business and consumer sides. Chief among them is the idea of human oversight and continued involvement in customer service.

When to make the transition from AI to human? 

This is something Xiaoming Liu, PhD, has thought a lot about. Liu works on biometrics and 3-D computer vision in his role as an MSU Foundation Professor and Anik K. and Nandita Jain Endowed Professor of Engineering at Michigan State University. His experience with AI models gives him a unique understanding on the limitations of AI.

Liu is a participant in the TP AI Think Tank, a cohort of researchers brought together by TP to better understand how the connection between EI and AI can translate to better experiences for both customers and customer experts. This research-driven exploration will provide organizations with a blueprint to ensure that any AI adoption is done intelligently and thoughtfully — keeping the human and emotional contexts front and center of all interactions. “There’s a real value that AI can bring. If we find the right problem set up, we can use the current technology for the good of society,” said Liu.

Liu suggests that the key is in understanding the technical boundaries of the AI system and knowing the right moment to transition from the AI system to human intervention.

“There will always be a certain percentage of cases the AI system cannot handle, and it's important for the AI system to be aware of its own limitations,” Liu said.

When the AI system reaches those limits, it should be able to seamlessly transition the interaction to a human agent, who can then address the more complex or unusual situations. It’s a fine line between switching too early and too late. Liu emphasizes that doing the switch too early would unnecessarily burden the human, while doing it too late could be dangerous if the AI system is making guesses it's not confident about. “Finding the right moment to transition is crucial for providing a good customer experience,” Liu said. 

As time passes and models get more and more robust, this transition process should gradually improve over time, with the AI system handling more and more routine tasks, while the human experts focus on the more complex or unusual cases the AI can’t handle. Liu is also excited about the development of Artificial General Intelligence (AGI), which are systems with more general intelligence, capable of flexible, adaptive and autonomous reasoning across a broad range of domains. AGI aims to have a wider understanding of things, as opposed to AI, which excels at specific tasks but can’t quite grasp nuance. 

Although humans will always be needed to oversee their efforts, both AI and AGI are technologies that show a lot of promise in being able to elevate the customer experience. “Sensing human emotions is very subtle and AI is not there yet. But business should keep an eye on AGI, because when AGI makes progress, a lot of downstream applications will open up,” said Liu.

AI and TP

As a leader in the industry, TP continues to be recognized as a global top employer — 97% of TP employees work in a certified Great Place to Work® or Best Place to Work®.

Part of TP’s secret to its success is how it supports customer experts with cutting-edge technology to streamline time-consuming processes and infuse customer interactions with emotional intelligence. According to the latest research from CMSWire and VKTR, organizations who say that they understand their customers well are much more likely to use AI extensively (33%) compared to those who understand their customers either moderately (7%) or poorly (3%). TP orchestrates, integrates, and manages AI-enabled and human-led solutions, leveraging a balanced approach that allows our TP experts to provide fast, reliable solutions while creating memorable, empathetic customer experiences.

Watch our interview with Dr. Xiaoming Liu and explore more expert insights in our YouTube playlist.