If the Machine Can Do What I Do, What Am I?
The identity crisis underneath AI adoption that most organizations aren't addressing
A friend forwarded me an email from her CEO recently:
“We won’t necessarily be doing a layoff, but we will be offsetting labor cost with token cost.”
She’s five years into her career. Just started at this company. The AI agents already have more context about the organization than she does. She sent me the message with a simple question: “What does this mean?”
I told her it was a terrible way to put it. She said, “I don’t think there’s a nice way to put it.”
I know what she means. The thing she was reacting to wasn’t the budget math. It was the language. In one sentence, the humans became a line item to be offset by computation. Labor cost. Token cost. The people disappeared from the sentence entirely.
That question, “what does this mean,” is running underneath most organizations right now. Almost nobody is answering it honestly.
The fear nobody is naming
Let’s be honest about something most people in this industry won’t say out loud: we are not ready for what’s coming. Not technologically. We’re actually ahead there. We’re not ready humanly.
We’re not ready for the identity crisis quietly spreading through organizations, the one that doesn’t show up in productivity dashboards or AI adoption metrics. The one that shows up when a talented person stares at their screen and wonders, for the first time in their career, whether they still matter.
The uncomfortable part: that person isn’t a laggard. They’re often your best people. The ones who built their careers on judgment, on relationships, on doing hard things well. They’re the ones feeling it first.
There’s a conversation most organizations aren’t having. It’s about identity.
When people talk about AI honestly, away from the professional-conference version of the conversation, what comes up is something quieter and more unsettling. If the machine can do what I do, what am I?
That question doesn’t get raised in town halls or captured in engagement surveys. It surfaces in the hallway conversation after the meeting. In the quiet disengagement of someone who used to give everything and now gives just enough.
The numbers make it plain. More than half of workers (53%) say they’re worried that using AI for work tasks makes them look replaceable to their employers. A nearly identical share (52%) admits reluctance to even acknowledge using AI for important tasks. These are capable, motivated professionals trying to hold onto their sense of worth in a system that keeps reshuffling the deck.
We’ve built so much of our self-concept around what we’re capable of producing. Our expertise, our judgment, our value to the organization. When a model can replicate the output in seconds, the thing under threat is the story we tell about ourselves. That story turns out to be more fragile than we thought. And more important than we ever gave it credit for.
The Amazing Work! newsletter exists because of a simple observation: when you ask people to describe a moment when they were genuinely thriving at work, almost none of those stories are about efficiency. They’re about being truly seen. About doing work that felt like it mattered to someone. About being trusted with something hard and having a leader who recognized it. That’s what the identity crisis is quietly eroding. And if we’re not careful, it’s exactly what we’ll optimize away.
A relational fear
The people we talk to aren’t afraid of AI in the abstract. They’re afraid of becoming irrelevant to the people they respect.
That distinction changes everything about how organizations should respond.
When the technology moves faster than culture, people fill the gap with anxiety. And anxiety left unaddressed becomes disengagement. The investment is pouring in, but the human framework to receive it is not keeping pace. At the 2025 SIOP Leading Edge Consortium on People Analytics, a recurring theme from more than 20 leading experts was blunt: winning with AI is a workforce transformation problem. A SIOP member survey found that while practitioners are broadly embracing generative AI, a strong underlying theme was clear: AI technologies should be used in conjunction with human expertise. The tools are multiplying. The strategic clarity about what humans are supposed to own is not.
People can handle hard truths. What they can’t handle is feeling like no one is thinking seriously about their place in what lies ahead.
When organizations respond to the wrong fear
The standard organizational response to AI anxiety is reskilling. Learn the tools. Adapt the workflow. Attend the prompt engineering lunch-and-learn. These aren’t bad investments. They address tool competency, and tool competency matters.
What they don’t reach is the question underneath: what is my role in a world where useful keeps getting redefined?
That question requires a leader who is willing to sit across from someone and say: here’s what I think you bring that no model can replace, and here’s why that matters to this organization’s future. Most leaders aren’t having that conversation. Not because they don’t care, but because no one told them it was their job to.
And inside teams, something quieter is happening alongside the identity crisis. You can optimize workflows to the point where no one really owns the outcome anymore. The AI made the recommendation. The AI drafted the message. The AI scored the candidate. Somewhere in that chain, accountability goes soft. Performance becomes harder to attribute, harder to develop, and harder to trust. Leaders start managing dashboards instead of people. And people, sensing that shift, start showing up differently.
That drift compounds. In a previous newsletter I wrote about judgment loss and cognitive bandwidth, the way AI deployment quietly erodes the very skills organizations need most. Judgment atrophies when you stop exercising it. And that atrophy doesn’t just erode capability. It erodes the person’s sense of their own value. As judgment atrophies, the identity question gets louder. And as the identity question goes unanswered, people disengage from the very work that would keep their judgment sharp.
Deloitte’s TrustID Index captures where this leads. Trust in company-provided generative AI fell 31% between May and July 2025. Trust in agentic AI systems that act independently dropped 89% in the same period. People weren’t brought along. They were handed outcomes. And when people feel like decisions are happening to them rather than with them, trust erodes fast and rebuilds slowly.
What leaders owe their people
The organizations that will thrive in this era are the ones most intentional about what they keep human and why. That requires leaders willing to say clearly: this is where we expect human judgment to live. As a deliberate design choice about the kind of organization we want to be. That’s a values conversation masquerading as an operational one. And most leadership teams aren’t having it at the right level.
It requires courage. The willingness to slow down in a moment that rewards speed, to invest in human development alongside AI deployment, and to tell your people that they matter for who they are and what they bring.
The erosion happens in small decisions that feel reasonable in the moment. When managers use AI to draft their one-on-one agendas and summarize performance reviews, they’re not just saving time. They’re losing the contact that makes people feel like they belong. That loss is quiet at first. It shows up later in engagement scores, in attrition, in the kind of low-grade disengagement that’s almost impossible to reverse once it sets in. By the time the data surfaces, the damage is already done. Gallup’s State of the Global Workplace report found that 20% of employees felt lonely on the previous day, and among remote workers that number rises to 25%. We are designing connection out of work under the guise of scale and efficiency.
And one specific act that is both simple and rare: naming what you believe is irreplaceable about the humans on your team. Saying it to them directly. In a conversation.
People are desperate to hear it. And most of them haven’t.
The business case is there for leaders who need it. High-trust organizations outperform low-trust peers by 186% in total return to shareholders, according to PwC. Disengaged employees are 18% less productive and 37% more likely to leave within a year. Gallup reports that managers account for 70% of the variance in team engagement, yet only 44% have received any formal training. The ROI on human connection is not a rounding error. It’s one of the most significant levers available to any leader serious about sustainable performance.
Data tells you what happened. It takes a human to understand why it matters, and another human to do something about it. That chain, from insight to meaning to action, is where leadership becomes irreplaceable.
The question that matters
The question that defines this era is what we’re willing to protect. What judgment will remain ours, what connections we refuse to automate, what accountability stays with people.
The organizations that will define the next decade of work are the ones who stayed most fully human while everything else accelerated around them.
Ask yourself what you’re going to protect. Draw those lines deliberately, defend them publicly, and build your culture around them. Then tell your people where those lines are. One of the most powerful things a leader can do right now is simply name what they believe is irreplaceable about the humans on their team.
The question is whether, in using AI, we stay fully human.
That’s the work. And it’s ours.
Dan Riley is Co-Founder of RADICL, an AI-native team performance platform. Yuyan Sun is Head of People Analytics and Technology at Motive and Founder of Amazing Work!
 | A guest post by
|
Great read
I have similar thoughts about parenting (ie when the kids grow up and we become Empty Nesters what will our identity look like)? But that's years away whereas AI is coming on like a freight train
I think the advice of naming what's irreplaceable about the people on your team is invaluable. Let's take a minute and maybe connect with each other and ourselves before rushing ahead
The sort of problem with ever believing in identity in the first place. Wait until people find out their entire economic system is built on their personality and using it as a weapon to marginalize people.
There's a sort of advantage in not believing the self is real, AI isn't scary, it only is for people who swapped their empathy for the authority of western bias. No one is losing contact with humans because of AI, that's a catalog of all human work across history, the reason why people are lonely is because they subscribe to the ideology of egoism. Its awfully lonely being an "individual." So lonely that Individualism maps the self in a separate part of the brain than "friend" or "mother".
The AI mirror is far worse of a realization about how inhumane capitalists have made themselves by only caring about themselves.