AI or Human: Which is a Better Listener?

I conducted a poll.

It was very informal and completely unscientific. I didn’t use any digital devices or spreadsheets, or even a pencil and paper. All I did was tally people’s reactions to the sight of my right hand wrapped in a bright purple cast, the result of an inopportune fall. It wasn’t very difficult. My poll produced a perfect score: 100% of the people who saw the cast responded by telling me about their experience with a cast or a hairline fracture; none expressed any empathy for my discomfort.

Chances are you’ve experienced the same when you showed up with conspicuous evidence of a medical treatment like a bandage or eye patch. You hear their point of view not yours.

Had I sought a reaction from an AI companion such as, Xiaoice, the responses would have been very different, or so Jamil Zaki, a professor of psychology at Stanford University, tells us. In a Wall Street Journal article based on his book, “Hope for Cynics: The Surprising Science of Human Goodness,” Professor Zaki reported that chatbots do “a better job than humans at making people feel seen and heard.” The reason, he explained, is that bots have “no personal experiences to share, no urgency to solve problems and no ego to protect, they focus entirely on the speaker. Their inherent limitations make them better listeners.”

That’s good news for AI and bad news for humans. Because no matter how powerful AI becomes, humans must communicate with humans. If that interpersonal exchange lacks empathy, if the communication is only one way—the point of view of the sender—there is no closing of the loop and no communication. The communication is likely to fail.

Of course, the stakes in business are much higher that the petty annoyance one might feel when the opposite party drones on and on about their similar medical experience.

The solution for all your business communications—whether in a meeting or team chat or a presentation—is for you to display empathy to the other party as effectively as a chatbot does. Professor Zaki tells us that ChatGPT has a “go-to recipe of ‘paraphrase, affirm, follow up.’”

For humans, it takes four similar steps:

1. Listen: Carefully take in your counterpart’s message and do so without formulating your response. The way to assure your focus is to subvocalize. Under your breath, say to yourself, “They’re saying/asking that…”
2. Paraphrase: Tell your counterpart what you heard them say. “You’re saying/asking…”
3. Confirm: Look for a head nod from your counterpart. If you do not get a head nod or worse, get a frown, go back to Step One and repeat until you get the head nod.
4. Respond: State your answer or position as it relates to your counterpart’s point of view.

Professor Zaki calls the effective responses of chatbots “LLMpathy” because they are driven by Large Language Models.

Human empathy is driven by an even stronger force: the brain.

In 1992, a team of Italian scientists published a seminal paper that reported on a study in which monkeys repeated behavior they observed. The scientists attributed this reaction to “mirror neurons,” a set of cells in the brain that produces empathic behavior. The phenomenon became known as “Monkey see, monkey do,” because it demonstrated empathic behavior. The most common form of empathy among humans is yawning. When one person in a group yawns, others yawn, too, in a contagious reaction.

When you demonstrate empathy to your counterpart in communication, their mirror neurons will impel your counterpart to respond to you affirmatively.

It’ll be contagious.