Death. Taxes. The speed of light. These are but some of the universal constants.
Of all the universal constants, though, change is perhaps the most pressing: stalking us daily at work, at home, and everywhere in between.
Given our depth of experience with change, one would think we’d be well equipped to tackle it—and to be fair, we are, or at least we are in terms of thinking strategically, finding new efficiencies, and, hopefully, updating our priors to suit the new environment in which we find ourselves.
That said, there’s a key component missing from that list.
The missing piece? People. Rarely do we discuss change from the perspective of the people experiencing it. Too often, we focus instead on urgent language like adapt or fall behind. Disrupt or be disrupted. Reinvent yourself. Reskill. Rebrand.
In the right contexts, that’s reasonable advice, but there’s a difference between understanding why change is necessary and the experience of that change, particularly when it’s a change that takes something from you.
And ever since the earliest days of the industrial revolution, one of the most destabilizing examples of this has been the threat change poses to hard-earned skills and craft.
Craft Under Threat
The Luddites are largely remembered as a group of yokels who simply hated technology, but the truth is far more complex. The industrial revolution turned the Luddites against specific machines not because the machines were new, but because those machines threatened their livelihoods, independence, and status.
Most of the Luddites were skilled textile workers whose craft mattered to them immensely. They’d spent years developing expertise within a formal system of apprenticeship and had a deep understanding of their work. As a result of their dedication and consumer demand, these skills had economic value. More importantly, however, this combination of factors granted these textile workers control over their work lives and a sense of identity within their communities.
That is, until the machines came along.
With textile production reorganized around machines over manual labor, skilled workers came increasingly under threat, ultimately leading to ways of work in which individual craftsmanship mattered less and less.
Sure, the Luddites may have been unusually willing to raise hell about putting a stop to this change of circumstances1, but despite their best efforts, the change they were reacting to didn’t disappear—it repeated. Industrialization continued to break complex crafts into smaller and smaller tasks, and it, along with other types of automation, continues to do so to this day.
In case no examples come to mind, think of the assembly line. This way of reorganizing work led to the production of goods at a scale that would have been unimaginable to earlier craftspeople, many of whom underwent this social transformation from highly skilled workers to meat appendages for highly repetitive, low-skill tasks.
Then there’s the mechanization of agriculture, the automation of factories, globalization, outsourcing, and software’s elimination of entire categories of clerical activity.
We’re seeing this again now, too, with artificial intelligence2 imposing on forms of knowledge work that, until very recently, seemed relatively insulated from that history.
The point is that the technologies may be different, but the experience is surprisingly familiar, leaving the displaced or to-be-displaced with a frightening pair of thoughts: I spent years learning how to do this. Does that still matter?
That question deserves more attention from anyone responsible for leading people through change—no matter the scale—because a threat to craft is rarely a threat to someone’s task list alone.
These changes threaten autonomy. They threaten belonging. They often fundamentally undermine competence, at least in the short term.
Autonomy, relatedness, and competence are the three psychological needs that underpin self-determination theory, a framework I’ve returned to before when thinking about meaningful work and craftsmanship.
In the workplace, I tend to think of them as something even simpler:
Autonomy. Belonging. Competence. The ABCs of finding your way at work.
Periods of upheaval, unfortunately—particularly when the upheaval is sudden or unending—have a habit of undercutting one’s sense of all three.
Autonomy: Do I still have a say in my own work?
Becoming a master of your craft traditionally offers you control. The more expertise you possess, the less you generally need to be told exactly how to perform every part of your job.
A craftsperson understands the materials. The tradeoffs. The exceptions. They have the hard-earned judgment necessary to outperform their lesser-experienced peers. Expertise creates discretion, or at least it can, which is one reason technological and organizational change can feel threatening even when someone’s employment itself is not immediately at risk. Change can alter not just what someone does, but how much authority they retain over doing it.
Here, too, we find a long history supporting this notion. Industrial systems have often improved output in part by moving decisions away from individual workers and into processes, and scientific management3 only pushes that idea further by studying work, determining the optimal method, standardizing it, and organizing workers around the system.
In these types of work environments, employees are told which technologies to use, how often to use them, and perhaps eventually judged on whether they’re using them enough. At that point, a technology that might otherwise have expanded human capability begins to feel like something being done to people instead.4
That distinction matters because what was originally an attempt to increase an individual’s autonomy by, say, permitting them to “focus on higher leverage tasks” ultimately removes some meaningful agency those individuals previously had over their work.
At times of change, then, the managerial question should not be limited only to whether people are using a new technology, but whether they are still experiencing meaningful agency over their work despite this new technology.
That said, no leader can preserve perfect autonomy during every change. Sometimes systems need standardization. Sometimes workflows genuinely improve when they’re redesigned. Sometimes employees will have to learn tools they would not personally have chosen. I get all of this; I’m a huge fan of standardization, efficient processes, and improved workflows. My career is riddled with examples of pursuing just that!
But when pushing for change myself, I always try to think carefully about the impact it will have on people. If the change in question is going to remove some amount of control from those with whom I work, it’s important to make sure those people experience an increase in control elsewhere, even if it’s in the implementation process for the change itself.
The people closest to the work are worth consulting about where automation helps and where it doesn’t precisely because they’re closest to the work. Even if that means they may have blind spots of their own, their insider knowledge is invaluable and should be the cornerstone of any conversation about change.
In other words, to ensure autonomy isn’t undercut during times of change, it’s incumbent on leadership to increase ownership over higher-order decisions when lower-order ones become automated.
Change creates opportunities for management, too. Some leaders will see disruption as an opportunity to consolidate control while employees are already on their heels; others will see an opportunity to create new forms of ownership and participation.
Change is an opportunity; the type of opportunity management sees in it is often very telling about the relationship they have with their reports.
And if circumstances are already taking autonomy away from your people, that is certainly not the time to unduly take more.
Competence: What happens when the thing you’re good at matters less?
The effect of change on competence is where craft under threat becomes especially personal.
Just think, for decades—if not since time immemorial—we’ve given people a particular kind of career advice: Learn valuable skills. Develop expertise. Build mastery and become excellent at something.
I’ve made versions of that argument myself.
The “craftsman mindset” is compelling precisely because competence can produce so many other things we want from work. Mastery creates pride. Scarce skills create leverage. Expertise produces career capital that eventually buys greater flexibility, interesting opportunities, and autonomy.
But—and this is a major but—there’s an uncomfortable implication to that advice.
What happens when someone does exactly what we told them to do—say, go to college and take out massive loans to get an in-demand degree—and then the world changes what their skill is worth?
Yeah, not great, and this has happened repeatedly.
The weaver becomes less economically valuable when a machine can produce cloth at dramatically greater scale. The factory craftsperson loses some of the importance of holistic knowledge when production is decomposed into standardized steps. The office worker becomes less necessary when software automates the process.
And here, now, we watch as experienced knowledge workers are left gobsmacked as artificial intelligence performs, in seconds, a task they spent years learning to do.
True, none of this means the machine can necessarily replace the person. Psychologically, on the other hand, replacement is not the only thing that matters. Competence is more than the possession of skills, it’s the experience of being capable, of knowing what you understand has value, and being able to look at a difficult situation and think, I know how to handle this.
When technological change undercuts that feeling, telling someone to simply “reskill” is an incomplete response. Of course people should learn new things, and of course skills evolve. No craft has ever been completely frozen in time.
But a leader who responds to someone’s professional disorientation with little more than “learn the new tool” risks sending a brutal message: What you spent years becoming good at no longer counts.
Yeesh.
The better response is to help people see both discontinuity and continuity, including the new opportunities that present themselves as a result of these changes.
Yes, parts of the work may become automated, but judgment still matters. Context and experience still matter, and knowing when an output is wrong still matters.
Knowing which problem to solve—perhaps the highest-order task available—still matters.
So, when competence is undermined, give your people harder problems. Give them responsibility, let them mentor, and let them learn. Their expertise is still valuable, it may just need new opportunities to demonstrate its flexibility across domains and new processes.
We can’t confuse the decline of one task with the disappearance of the person who used to perform it. That distinction is important because our rhetoric around change often blurs it.
When someone hears they’re now a “10x employee” and that “one person can now do the work of an entire team,” there are nine other people on that hypothetical team receiving that message, too.
Leaders should be thoughtful about what that message does to someone’s sense of competence.
Belonging: Am I part of whatever comes next?
Then there’s the question beneath almost every major period of workplace upheaval: Is there still a place for me here?
This is where change becomes bigger than craft. People can be led to ask themselves this question as the result of a merger or acquisition, a reorg, a new leader.
And layoffs will certainly have people asking themselves whether there’s still a place for them here. Even if they survived a first wave, what if there’s another?
The circumstances can vary enormously, but the underlying question doesn’t.
People want to know whether they remain part of the group and whether the group still has a future for people like them. This is why I prefer the word belonging in workplace conversations even though self-determination theory traditionally uses relatedness.
Belonging gets directly at the organizational experience and, even more importantly in my view, ties into something deeply human: the need to belong.
And whether organizations realize it, they’re constantly answering the question of whether their members are still wanted.
Think about the language of AI transformation today. We hear a lot about the future of work, AI-first organizations, smaller teams accomplishing more, automated agents, flattened organizations, efficiency, and headcount leverage.
It’d be hard to argue that employees are acting irrationally when they listen to all of that and wonder which category they belong to: are they the future or the inefficiency being removed from it?
Now, this doesn’t mean leaders should promise permanence. To do so would be deeply irresponsible when one considers that despite any workplace’s best intentions, roles will sometimes disappear. Sometimes organizations genuinely need fewer people doing a certain kind of work. Sometimes external circumstances force painful decisions that no good manager can prevent.
Belonging doesn’t require dishonesty, no, it requires people to believe that they’re being treated as participants in change rather than as inputs to be optimized.
There’s an enormous difference between nothing about your job will ever change and we want you to help us build what comes next. The first may be impossible to promise, but the second is often entirely within a leader’s control.
What change takes away—and what leaders can put back
This, I think, is where the ABC framework for leading through change becomes most practical. Change will sometimes diminish autonomy, belonging, or competence. That’s inevitable. The mistake is assuming that because the loss is unavoidable, nothing needs to be done about it.
To put it in terms even the most data-only executives would appreciate, imagine instead that leaders treated these needs as a kind of psychological balance sheet.
If a change takes something away, ask where you can put something back. For example, if automation removes discretion from part of someone’s work, can you increase their ownership elsewhere? If a new technology makes an existing skill less central, can you give that person opportunities to develop or demonstrate competence in a different way? If a reorganization creates uncertainty about someone’s place in the organization, can you increase communication, participation, investment, and connection?
The goal in doing this isn’t to ensure that nobody ever loses anything. That would make meaningful change impossible. The goal is to recognize the loss and, just as importantly, avoid compounding it.
If someone is already questioning their competence, that’s not the time to start micromanaging them. If people are already unsure whether they belong, don’t communicate with them exclusively through the language of efficiency. If a transformation is already changing someone’s work in ways they can’t control, don’t strip away the autonomy they still enjoy.
In other words, don’t stack deprivation on top of deprivation.
This principle applies far beyond AI, with COVID providing an obvious recent example.
During the pandemic, people lost enormous amounts of control over their working lives almost overnight: teams lost familiar ways of connecting, and employees had to learn unfamiliar technologies and new ways of doing jobs they already knew how to do.
Some managers responded to that upheaval by increasing flexibility, trust, communication, and compassion, while others responded to the loss of physical visibility with more surveillance and more control.
The external event was the same, but the managerial response was not—and that’s an important distinction. Leaders can’t control every disruption, but they do control whether their response restores something that’s been lost or takes even more from their people during challenging times.
Resistance is information
This also changes how we might think about “resistance to change.”
Yes, people can be stubborn, but “resistance” describes behavior without explaining its cause. This is yet another reason the Luddites remain such a useful cautionary tale.
Here we are, two centuries later, their name synonymous with all things technologically averse, but, like we highlighted earlier, their objection wasn’t simply machine bad. The behavior was resistance, but the cause of that resistance had its origins in the fact that their world of work was changing. Their expertise was being devalued. Their autonomy was being undermined. Their economic and social standing became less certain.
Talk about a hit to autonomy, belonging, and competence.
Some may argue this doesn’t justify everything they did—and they’re welcome to that opinion—but no matter how you look at it, these facts do make their behavior more intelligible.
The lesson here is that modern leaders should be careful not to make the same interpretive mistake: when people resist a change, ask what they believe they’re losing. Autonomy? Belonging? Competence? Some combination of the above?
The answer to those questions may not mean the change should stop, but it may tell you a great deal about how the change needs to be led.
The business case for self-determination theory is stronger than it looks
There’s another reason asking ourselves these questions as leaders matters: people who feel safe, capable, and valued have much better incentives to help an organization change.
Let’s take AI as an example again. Suppose I discover that a tool can automate thirty percent of my current work. If I trust my leaders, believe my expertise is valued, feel I have a future in the organization, and expect to have some say in what happens to the time that automation frees up, that discovery is exciting.
But in an organization where leaders are indifferent toward reductions in headcount, make unilateral decisions about work, treat employees as a line on an expense sheet, and describe AI mostly in terms of how many people will no longer be needed—that same discovery is no longer exciting. It’s a threat.
If I, or anyone, really, were in the latter organization, keeping that tool to myself wouldn’t necessarily be resistance to a full-throated embrace of technology. It might just be rational behavior.
All of this leads to a strange irony: the organizations most capable of benefiting from automation may be the ones where employees feel safest helping automate their own work.
In this way, we can see trust is a prerequisite for transformation.
And with autonomy, belonging, and competence our fundamentals to building trust, they should be treated as prerequisites to transformation, too.
The ABCs
Periods of upheaval come in many forms.
Some are global, some are organizational, and others, still, are intensely local. The scale of the change can, well, change, but the human questions are remarkably consistent.
To get ahead and stay ahead when leading people through change, then, it’s critical that we ask ourselves the below.
Autonomy: Does this leave people with more or less meaningful agency over their work?
Belonging: Does this give them more or less reason to believe they have a valued place here?
Competence: Does this help them experience themselves as capable and useful, or does it make something they spent years learning suddenly feel irrelevant?
The answer may not always be positive. Change sometimes takes more than it gives, and leadership isn’t the art of preventing that; it’s the practice of understanding what change is asking people to give up, deciding which losses are truly necessary, and being intentional about what can be restored elsewhere.
Because while the history of work is in many ways a history of changing tools, systems, and forms of craft, the people doing the work have remained consistent in what they need. People need agency. They need connection, control, and a sense of mastery—our ABCs of autonomy, belonging, and competence.
Work will change. It always has.
Our obligation to the people doing that work should evolve with it.
Join the Conversation in the Comments
Think of a change you were asked to navigate at work: what did your leaders do well, and where did they lose you?
Which matters most to you during a period of change—autonomy, belonging, or competence—and why?
What is one thing leaders consistently misunderstand about resistance to change?
What started as dust-ups outside factory walls slowly turned into carefully organized property destruction and, later, riots, beatings, and hangings. The Luddites may have failed in their ultimate goal of putting a stop to the automation of their craft, but the protections they succeeded in securing led to the creation of the world’s first unions. For more on the full history of the Luddites, I recommend reading the fascinating book Blood in the Machine: The Origins of the Rebellion Against Big Tech by Brian Merchant.
By this, I don’t mean to suggest that AI is doing this on its own. On the contrary, I’ve written in the past about how technologies do not advance themselves. There are always people behind these initiatives. It’s these individuals whose motivations and credibility require scrutiny just as much as—if not more than—their technologies themselves.
Or, in the event that management is insufficiently scientific to pull this off on their own, they hire consultants whose jobs, interestingly, may be among some of those that soon find their own sense of autonomy, belonging, and competence turned on its head in the age of AI.
When humans-in-the-loop go from individuals who control the work to individuals whose role it is to perform far simpler, less skilled tasks, they become what is known as a reverse centaur. I suspect I’ll have much more to say about reverse centaurism once I have the chance to read Cory Doctorow’s The Reverse Centaur’s Guide to Life After AI: How to Think About Artificial Intelligence—Before It’s Too Late.