Why Emotional Intelligence Matters More in the Age of AI

Rich Hua spent years at Amazon helping bring emotional intelligence into a culture famous for analytical rigor. His interest in EQ, though, started much closer to home. He describes his younger self as a “genius robot”—highly analytical, successful by conventional measures, and emotionally underdeveloped. Marriage and fatherhood forced him to rethink that model.

I spoke with Rich about learning emotional intelligence, teaching it at scale, and why increasingly empathic AI creates both opportunities and risks.

Paul Zak: You describe your younger self as a “genius robot.” What convinced you that intelligence alone wasn’t enough?

Rich Hua: Being analytical worked really well for me for a long time. I got good grades, studied engineering, and even got a perfect score on the SAT. Then I got married and found all of that grossly insufficient for being the caring, empathetic, loving husband I wanted to be.

Having kids made it even clearer. My default view was that unpleasant emotions—anger, anxiety, disappointment, insecurity—were distractions. I had to learn to acknowledge emotions rather than discard them. I started sitting with discomfort instead of immediately trying to make it disappear. It was like working out a muscle. At first it was incredibly uncomfortable, but over time I got better at it.

PZ: You eventually scaled emotional intelligence training at Amazon to more than a million people. What did you learn about whether EQ can actually be taught?

RH: It absolutely can be taught, but people start from different places.

Someone like me needs to open up more and become aware of emotions. Someone at the other extreme may absorb everyone else’s emotions and need stronger boundaries.

One pattern I saw among highly analytical leaders was that they were often very good at getting things done. They tended to be high on self-management and relationship management, so they could push themselves and others to solve problems and deliver results. But they were often less developed in self-awareness and social awareness.

That creates problems. They go, go, go without recognizing their own stress, anxiety, or exhaustion until they burn out. And when persuading other people, they sometimes think, “If I just give them more data, surely they’ll agree.”

But leadership requires winning hearts and minds, not minds and minds.

PZ: Companies spend enormous amounts on leadership training that produces little lasting behavioral change. How did you know the Amazon programs were working?

RH: The dirty little secret of training is that most of it has no effect on behavior.

Training alone isn’t enough. You need three things: ability, motivation, and opportunity.

Ability means learning and practicing the skill. Motivation means creating an environment where the desired behavior is expected, supported, rewarded, and modeled by leaders. If you tell people to be empathetic but promote people who behave like jerks, the training isn’t going to matter.

Opportunity means giving people room to practice, make mistakes, get feedback, and improve.

At Amazon, the EQ movement grew from me giving talks here and there to a community of practice that eventually numbered 70,000 people. We saw leaders change what they actually did—for example, introducing gratitude practices into their team meetings. In surveys, more than 90% strongly agreed that EQ training helped them do their jobs better, improved workplace relationships, and was something they would recommend to colleagues.

That’s much more meaningful than simply asking whether someone enjoyed a class.

PZ: You also advised the Alexa Plus team on empathy and emotional intelligence. What does it mean for an AI to be empathic when it doesn’t experience emotions?

RH: AI can recognize what someone is experiencing and respond appropriately. It can say, in effect, “I understand what you’re going through, and I want to support you.”

But we were very thoughtful about not creating over-reliance. We wanted Alexa to support human flourishing, not become so engaging that people wave off other humans.

That’s the danger. Humans are messy. In contrast, AI is always available. It doesn’t have a bad day. You don’t have to ask how it is doing.

If we start offloading social and emotional connection to AI, our social skills can atrophy. Friction with other humans is sometimes unpleasant, but learning to navigate that friction develops empathy, adaptability, and social skills.

PZ: Make the case against your own argument. If AI becomes exceptionally good at reading people, resolving conflict, and communicating empathically, which human EQ skills become less valuable?

RH: I divide interactions into two buckets: transactional and relational.

AI can replace a lot of both IQ and EQ in transactional situations. Customer service is already moving in this direction. If I need a refund, I may prefer an AI that understands exactly what I need and solves the problem immediately.

But relational situations are different. If you’re dealing with trauma, the death of a child, a separation from your partner, or another profound experience, do you really want only an AI trained on good frameworks?

There are things AI cannot genuinely embody: care, lived experience, pain, meaning, purpose, morality, and sacrifice.

Sacrifice is uniquely weighty. AI makes no sacrifice. A human does. Giving up something important for another person is one of the deepest ways we show we care.

So as AI gets better, the dividing line becomes clearer. Machines will become extraordinary at transacting and engaging with emotional intelligence. But the deepest human relationships require more than the performance of empathy.

They require another human being.

FAQs

1. Can emotional intelligence actually be taught?

Yes. Emotional intelligence can be developed through practice, self-awareness, and feedback. However, effective training requires more than instruction. Organizations also need to create environments that encourage, reward, and provide opportunities to practice emotionally intelligent behaviors.

2. Why do so many emotional intelligence training programs fail to change behavior?

Many training programs focus on teaching skills without addressing the environment in which those skills are used. Lasting behavioral change requires three things: ability, motivation, and opportunity. Employees need to learn new skills, see those behaviors modeled and rewarded, and have opportunities to practice them.

3. Can artificial intelligence genuinely demonstrate empathy?

AI can recognize emotional cues and respond in ways that appear empathetic, but it does not experience emotions or care in the way humans do. This distinction becomes especially important in relationships involving vulnerability, lived experience, and personal sacrifice.

4. Will AI make human emotional intelligence less important?

AI may increasingly handle transactional interactions that benefit from emotionally intelligent communication, such as customer service. However, deeper relationships still require human qualities like genuine care, shared experience, and sacrifice. As AI becomes more capable, developing and maintaining these human skills may become even more important.