No, AI won’t steal your career. Trust me, I’m a sociologist.

The year is 1999, and you’re a librarian. For decades, your profession’s main job has been information retrieval and interpretation. If someone in your community had any kind of question---What’s the boiling point of water at high altitudes? Where’s the nearest dentist?---you would comb through various sources of information, synthesize your findings, and provide an answer.

Now, the walls are closing in: a new zero-cost-to-consumer website called Google allows anyone with an internet connection to conduct information search, which is about 90% of your job. Google is fast, convenient, and seemingly omnipotent. Meanwhile, your skills (navigating card catalogs, databases, and academic journals) feel increasingly undervalued. There’s a growing sense that librarians are doomed.

Fast-forward to the year 2025, and librarians are still very much a part of our society. How? Why didn’t the library profession disappear if 90% of their job was automated away?

You’ve probably heard lots of fearmongering lately, usually along the lines of “countless workers will soon be replaced” as AI continues to automate tasks. Headlines warn of impending obsolescence, with countless careers on the brink of extinction.

Here’s the problem. The modern workforce is increasingly dominated by the professional and technical sector, which makes up nearly 3 in 5 American workers (and counting). Why does that matter? Because all the hand-wringing panic attacks about "THE ROBOTS ARE COMING FOR OUR JOBS!" completely miss the point. Professions don’t just roll over and die and when new technology appears. In fact, they respond to changing circumstances in some unique but predictable ways. The fearmongering about mass replacement by AI stems from a fundamental misconception about what professions really are and how they respond to technology.

Many people assume that a profession is, fundamentally, a collection of tasks, or a type of work someone does. Lawyers learn and perform lawyerly tasks, like writing contracts. Doctors learn and perform the tasks of being a doctor. So if those tasks can be automated, those professionals will be out of a job.

Though this seems straightforward, it actually misses an important detail: over time, the specific tasks that make up a profession change a lot.

Take medicine. A doctor from the 1920s would have no idea what today’s physicians are doing. The stuff that once defined the job---calculating oxygen levels and other vital signs, manually analyzing tissue samples under a microscope---has all been automated away.

And yet, doctors are still here. In fact, they’re busier than ever, just with completely different tasks. A century ago, specialties like Oncology, Immunology, or Emergency Medicine didn’t even exist. Half of modern medicine would look like science fiction to their predecessors: robotic surgery! CT scans! organ transplants from the dead!

So if a profession isn’t just a bundle of tasks, then we can’t define them based on the tasks they happen to be doing right now. But if that’s true, then…

What even is a profession?

Sociologist Andrew Abbott argued that professions are like animal species: constantly adapting, defending their turf, and sometimes getting pushed out by bigger, meaner competitors. This isn't a predictable or preordained process, because professions don’t just passively exist within an ecosystem. Rather, they actively influence and transform their environments. They stake claims, fend off threats, and, when it suits them, straight-up steal from each other. They may rely on one another for knowledge and skills, but make no mistake---this is a fight to the death for resources, recognition, and authority.

The key to Abbott’s theory is the idea of jurisdiction. Professions aren’t just defined by what they do, but by their ability to convince the rest of us that only they can do it. These ideas are reflected in our licensing bodies, educational systems, and regulatory frameworks. For example, physicians are the only profession that we allow to perform surgery on humans, partly due to regulations, partly due to our trust in the field of medicine, and partly due to the profession’s political maneuvering.

So what happens when new technology rolls in? It tends to deskill tasks, making them easier, faster, and accessible to less specialized workers. This isn’t new. Assembly lines deskilled craftsmen, turning them into machine operators. Google deskilled information search, making librarians less essential as the gatekeepers of knowledge. Once anyone could do a task, it was no longer theirs to own.

This is what’s at stake with AI. It’s not just about automation; it’s about who gets to keep their monopoly on expertise.

That said, technology also creates new types of work. For example, 60% of jobs in 2018 did not exist in 1940. The good news? New tasks mean new jurisdiction is up for grabs.

Imagine if a large volcanic island suddenly emerged from the ocean. Now imagine the geopolitical chaos as countries scramble to claim it. The same is true with professions: as new tasks appear, it’s not always clear which profession will claim ownership. And just like countries redrawing borders, professions are constantly battling over who gets to claim which tasks.

It’s not always pretty. Consider nurse practitioners, who have seen a dramatically expanding scope of practice since the 1960’s despite the best efforts of physicians groups to defend against it. Fast forward to today, and NPs in many states can diagnose illnesses, prescribe meds, and even perform procedures that used to be the jurisdiction of medicine alone. What was once the sacred, untouchable domain of physicians is now contested.

This isn’t to say that AI won’t have major impacts on the workforce. It will. Some researchers have estimated that large language models will affect at least 10% of work tasks for four out of five U.S. workers, while approximately one in five workers may see half or more of their tasks impacted.

So yeah, AI is going to change things. But the key point here is that jobs aren’t static. They evolve, stretch, and adapt, whether the people in them like it or not. Instead of panicking about AI “stealing” jobs, the real question is: How will your profession stake its claim in this new landscape?

So if AI is creeping in, what should you do?

If you have a specialized skill, listen up. Jurisdiction over tasks is not something that professions inherently possess; rather, it is something that society grants through cultural processes like laws, norms, and trust. This means that a profession’s claim to a specific domain of work isn’t an objective truth but rather a negotiated agreement shaped by the cultural, social, and institutional frameworks we’ve built over time. In essence, who gets jurisdiction (and how much authority they have within that jurisdiction) is determined by a society’s collective understanding of expertise, competence, and necessity. This is good news for you because it means that your profession can actively fight for survival.

In general, professions respond to these deskilling technologies through a combination of resistance, adaptation, and redefinition of their roles. The most successful professions will be those who can defend their jurisdiction against encroachment, or those that can grow into new jurisdictions when tasks become redundant. The professions that maintain exclusive control over important tasks will be able to maintain their status, autonomy, and economic opportunity.

So, the first move for any profession under siege? Defend. Lock the doors and make sure no one else gets a key. This is why we have regulatory boards, licensing bodies, and professional associations. They’ll say it’s for quality assurance, which is partly true---but let’s be real, they also exist to keep the amateurs (and now, the robots) from encroaching on the good jobs. Many professions have protected tasks, and by framing the use of technology as part of these protected tasks, they ensure that automation tools could only be used under their supervision.

This is why, when English lawyers were faced with tools that could automate tasks like document analysis, they insisted that these tasks remain within their control. Instead of fighting automation outright, they pulled a power move: they made sure only they could use it. How? They positioned themselves as the only ones qualified to safely oversee or perform tasks involving the automation technology. This way, AI wasn’t replacing them---it was just making their jobs easier, while still keeping non-lawyers (and robots) out of the game.

This approach often goes hand in hand with making ethical and safety-based claims about the dangers of removing humans from the equation. Highlighting risks, like the potential for AI-driven errors or the lack of empathy in decision-making, can effectively underline the continuing importance of human expertise.

However, there’s a cautionary tale here: the people best positioned to adapt to new technology are often the last ones to take it seriously. That’s the “paradox of expertise”: thinking you’re too good to be replaced until it’s too late. When you’re an expert on a topic, it’s easy to underestimate the usefulness of new technology just because it seems like a janky, inferior version of what you do.

Librarians learned this the hard way. When the Internet came along, they dismissed early search engines as chaotic, unreliable, and incapable of delivering the kind of carefully curated knowledge they specialized in. Because of this, they struggled to appreciate the value that regular people would find in Google and how it would transform the information economy. (See also: the writers and artist of today endlessly making fun of AI slop)

Over time, librarians got the memo: adapt or die. They stopped seeing themselves as human search engines and started leaning into what actually made them valuable---connecting people to the right information. Instead of clinging to the old ways, they figured out how to work with the tech: curating resources, creating specialized guides, and even building recommendation algorithms. By embracing this new identity, librarians reclaimed relevance and demonstrated the ongoing value of their expertise.

The lesson? Expertise isn’t about clutching onto old tasks. It’s about evolving into the next version of yourself before someone else does it for you.

This isn’t just about AI. It’s about making sure you---and your profession---stay in the game. Want to stay relevant? Start shaping the narrative. Write the op-eds. Join the professional orgs. Get loud in the policy meetings. Because the people who define the rules are the ones who get to keep playing.

So no, AI won’t steal your career. But if you’re not paying attention, another profession might.