empathy at work

This week Meta notified approximately 8,000 employees — roughly 10% of its workforce — that their positions were being eliminated, part of a sweeping restructuring the company has framed as necessary to fund its push into artificial intelligence. At the same time, 7,000 workers are being redirected into newly created AI-focused teams. The stated rationale is efficiency, investment, and building an AI-first organization.

I feel for the people who lost their jobs, and the communities who may be impacted by that large number of people being out of work. It it also gives me pause about how the company is redesigning work, with particular attention on human skills such as empathy, out of the work itself.

What Gets MIssed When Work is Redesigned Around Tasks Alone

When companies restructure around AI, the decisions are typically made at the level of structure and budget: headcount targets, function eliminations, and capital redirected from people to infrastructure. What is often not captured are the human skills embedded inside the work itself: the employee who senses what a customer needs before they say it, who hears the unvoiced question, and who finds the right words at the right moment to turn a frustrated caller into a loyal one.

Those human skills do not appear on a process map or show up in a task description. When the function is eliminated, they walk out the door with the people who held them, and may gradually be designed out of the jobs of the people who remain. Nobody budgeted for that loss. But the customer feels it fast.

What AI Cannot Do

We are humans before we are customers, before we are employees, before we are stakeholders. Emotions are our native language and they predate every other language we have ever developed. Long before human civilization developed words, people communicated danger, grief, and belonging through what they felt and what they read in others. As infants we communicate fear, hunger, comfort, and joy long before we have words for any of them, and the people around us learn to read those signals with extraordinary precision. That capacity does not leave us when language arrives. That is the foundation of human connection itself. AI was built on language. It has no access to the layer of human experience that exists before words form.

AI can define empathy, explain the research behind it, and describe how it functions in organizational settings. What it cannot do is feel it, and what it cannot do is intuit what to do about it in the specific, contextual, human moment when it matters. A leader sitting with a customer who feels dismissed, an employee who is frightened, or a partner who feels squeezed, is navigating something no algorithm can fully read. The human skills required in that moment are developed over time, through experience and attention and they are what no model can learn to feel.

The Cost of Getting It Wrong

The dominant framework sorted skills into two categories: technical skills (formerly called hard skills), considered measurable and therefore serious, and human skills (formerly called soft skills), everything else. What could be measured became synonymous with what mattered, and everything else was diminished. For anyone whose greatest strengths lived in the human rather than the technical domain, the message from organizations was that those strengths did not carry the same value. Contemporary management research has consistently revealed the importance of human skills such as empathy, communication, collaboration, emotional regulation, and judgment. Peter Drucker saw this clearly. "Management is a Liberal Art," he wrote in 1989, liberal because it deals with knowledge, self-knowledge, wisdom, and leadership; art because it deals with practice and application. Managers, he argued, must draw on psychology, philosophy, economics, ethics, and the humanities, and focus that knowledge on effectiveness and results.

We have devalued human wisdom and skills before. We know what it costs. The losses surface in the numbers, in eroding relationships, in customer and employee experience, and in the partnerships that dissolve when people no longer feel respected. With the urgency and scale of AI adoption, we risk repeating that devaluation faster than we can recognize it. Which is precisely why the signals already present in every organization deserve more attention, not less.

Named or Not, Emotions Are in the Room

Emotions are present in every organization, driving behavior, shaping how customers decide, how employees perform, and how vendor relationships hold or fracture under pressure. Leaders who cannot read them are making decisions with incomplete information. Warning lights do not stop flashing because you tape over them, and an organization that has trained itself to ignore emotional signals is not operating without that data. It is simply operating without access to it.

The Dashboard Every Leader Needs

Think of emotions as dashboard indicator lights for human needs. When needs are met, the lights are green: trust, engagement, loyalty, collaboration. When needs go unmet, warning lights flash: frustration, disengagement, churn, conflict. The people in your organization who can see the world through another person's eyes, anticipate needs before they are voiced, and find the right words at the right moment are your most valuable readers of that dashboard. Empathy is what allows a leader to see which needs are being met and do more of that, and to see which needs are not being met and work to address them.

Customers need to feel heard, respected, and reliably supported. Employees need clarity, psychological safety, and the sense that their contribution matters. Vendors and partners need to know that the relationship is fair and that their expertise is trusted. When these needs are met, the results are visible: loyalty, engagement, partnership, advocacy. When they go unmet, the costs accumulate in ways that rarely appear on the original task audit with churn, turnover, reputational damage, and the slow erosion of the organizational trust that makes execution possible. Leaders who dismiss emotions as irrelevant are missing the data they need to sustain loyalty with customers, build trust with employees, and maintain strong partnerships with vendors.

Empathy in Practice

Here are three practices to sharpen individual and organizational empathy:

• Ask people how they are feeling, and mean it. The people in your organization possess information that will not appear in any report, and the only way to access it is to create the conditions in which they feel safe enough to share it.

• Listen carefully to those who are putting emotions into words. They are not being difficult. They are giving you access to the organizational intelligence you need to make better decisions.

• As you consider if, how, and when AI tools are integrated into your organization, take a full accounting of the human skills associated with each role, not to automate them, but to understand the what the organization depends on and protect it accordingly.

Technical skills and human skills are interdependent. A strategy without the trust to execute it is a document. A financial model without the judgment to interpret it is a spreadsheet. An AI implementation without the empathy to understand what customers, employees, and partners actually need from it is an expensive source of new problems.

When technology serves human effectiveness, relationships, and organizational health, it fulfills its purpose. That is the standard worth holding: for every tool you adopt, every role you redesign, and every decision you make about how work gets done. Empathy is what keeps that standard visible.