What AI Anxiety Really Means: The Fight to Prove Your Value in an Automated World

“The past is never dead. It’s not even past.” — William Faulkner

What AI Anxiety Really Means: The Fight to Prove Your Value in an Automated World

Learning the new tool matters. It does not, by itself, guarantee a place in the organization the tool reshapes.

When an AI rollout makes you more productive, what has it proved: that you are more valuable, or that fewer people are needed for your current work?

That question is the sharpest part of Night King, a Hong Kong film set in 2012 Tsim Sha Tsui East. Using its nightclub as a lens, I want to separate learning a new tool from securing a place in the value chain that tool rearranges.

The surprise is that the nightclub’s transformation works. Performance improves. The new system looks modern. That apparent success is what makes the story unsettling.

The most frightening part of change is not being replaced by it; it is helping it replace you.

Run this query against one workflow you are automating: If this task becomes nearly free, who owns the next decision? If the answer is not you, tool fluency may be necessary, but it is not career insurance.

The makeover succeeds before its cost becomes visible

On the surface, Night King is about a once-thriving nightclub, East Sun, being acquired. Its longtime manager, Huan-ge, has spent decades in the night scene. Then his former wife, V-jie, arrives as the new CEO.

Former spouses become management opponents. A nightclub becomes a workplace. The workplace becomes a pressure chamber.

V-jie pushes Huan-ge and the female host team to transform how they work. Results begin to improve. Yet the improvement sits inside a larger capital-driven reshuffle. By the end, Huan-ge and V-jie move from fighting each other to working together to try to preserve East Sun.

That plot matters because it refuses the easy version of a workplace story: cruel owner, mistreated workers, righteous revolt, defeated villain. The new system produces visible gains. Change is not presented as obviously incompetent or pointless.

The danger arrives later: success at the new process can make the old role easier to remove.

Old-world value	New-world value	What can disappear in the translation
Personal relationships	Measured output	Trust built over years
Reading people and the room	Process compliance	Context that does not fit a spreadsheet
Loyalty and familiarity	KPI performance	The reason regulars return
Informal judgment	Standardization	The local knowledge behind good judgment
Experience	Efficiency	The role that experience once justified

The table is the film’s argument in miniature. Neither column is pure virtue. Relationship-based systems can be opaque and binding; metric-led systems can be more legible and more efficient. The problem begins when the organization treats what it can count as the whole of what creates value.

AI anxiety is often anxiety about the value chain

The familiar advice is simple: learn AI or AI will replace you.

I think that advice is incomplete. Learning the tool is sensible. Refusing useful tools out of principle is usually a poor strategy. But “become more efficient” does not answer the question that follows: who captures the benefit of that efficiency?

Suppose an organization can now do with two people what previously took ten. The other eight may have learned the same tools perfectly. Their competence does not recreate the previous number of roles.

This is why “transformation” and “progress” should not be treated as synonyms. Transformation means change. Whether the change leaves you with more agency, more bargaining power, or a role closer to consequential decisions is a separate question.

A software engineer can automate implementation work and still become easier to substitute if their contribution remains defined as producing implementation. The safer position is not mystical. It is closer to deciding what should be built, validating whether it worked, understanding the constraints the model cannot see, and carrying responsibility when the system is wrong.

That work can also be automated in pieces. Nothing grants permanent immunity. But it changes the question from “Can I use the tool?” to “What judgment becomes more important because the tool exists?”

Huan-ge and V-jie disagree about what value looks like

The former couple are more than a romantic setup. They are competing management philosophies.

Huan-ge stands for experience, relationships, favors, social reading, and the old rules of the trade. V-jie stands for management, systems, commercial discipline, performance data, and transformation.

Their conflict is not simply man versus woman, former husband versus former wife, or old versus new. It is a collision between two incomplete operating models.

Huan-ge’s methods once made him successful because the world around him needed those methods. That is why his situation lands harder than a standard failure story. He is not incompetent. He is attached to a version of competence whose market has changed.

Decades of experience can become decades of path dependence when the game changes. Past experience counts as experience only when the future still needs it.

The film’s title carries the same idea. A night king is a king only because a particular kingdom exists. The person may remain capable. The kingdom may simply stop needing that kind of ruler.

Metrics change the ruler, not the person

The nightclub is a useful setting because the contrast is so exposed.

The old world says: this customer knows me. This person has been with us for decades. We know the atmosphere, the regulars, and the unspoken rules.

The new world asks: what does this customer contribute? What measurable value does this employee create? Can that value be put in a dashboard?

People have not changed. The ruler used to measure them has.

That is the film’s critique of professional management when it becomes too certain of itself. KPIs have a legitimate purpose. They make some tradeoffs visible and can expose wishful thinking disguised as tradition. But when a manager assumes every important problem can be solved through KPIs, the KPI becomes part of the problem.

What is measurable gets optimized. What is hard to measure is easily neglected: trust, atmosphere, emotional labor, shared history, and the culture that made customers or colleagues stay. An organization can become more efficient while losing the thing that made it worth choosing.

The counter-case deserves to be stated plainly. Informal relationships can conceal favoritism, resist accountability, and make an organization impossible to scale. A metric is sometimes the first honest description of a problem. The answer is not to romanticize the old system. It is to ask which valuable behavior the metric fails to observe, and who pays when it disappears.

The real opponent is the self that used to win

It is tempting to make capital the sole villain. The film is more interesting than that.

Huan-ge is loyal to an old world because that world gave him identity, dignity, status, friendship, love, and the feeling of being useful. He is not only defending a nightclub. He is defending the evidence that he mattered.

That is a familiar trap in engineering too. We keep reaching for the technique that made us successful: the language we know best, the architecture we once rescued, the process that protected us from a previous failure. Sometimes that is expertise. Sometimes it is a dependency on being the person who knows the old answer.

The lead performance fits this conflict because its comic energy is built around verbal defiance. The jokes can make the world, bosses, men, women, and the speaker himself look ridiculous. Under that hard talk is resignation: the world is absurd, and one still has to live in it. That is a more adult comic posture than winning.

Night King does not say the old era was good and the new era is good. The old one binds people through relationships and opacity. The new one binds them through performance measures and data. Absurdity does not vanish with modernization; it gets a newer interface.

Don’t reject the tool. Locate your leverage.

The practical response to AI anxiety is neither blind enthusiasm nor nostalgic refusal.

Learn the tool. Use it enough to see where it is reliable, where it fails, and which parts of your work it makes cheaper. Then inspect the organizational consequence of that cheaper work:

• Does the saved time create room for better judgment, or merely reduce headcount pressure?
• Do you own a decision, a customer relationship, a system boundary, or an outcome that becomes more consequential after automation?
• Are you making hidden context visible before someone replaces it with a convenient but incomplete metric?

This is the disagreement I am willing to defend: “Learn AI” is weak career advice unless it is paired with a plan for where you sit after AI changes the value chain.

There are limits to this framing. Some teams will use automation to expand what they can build rather than shrink roles. Some individual contributors will gain real leverage simply by becoming unusually good at directing and evaluating AI systems. The point is not that efficiency always destroys jobs. It is that efficiency alone does not promise anyone a continuing place.

The supplied film notes describe a roughly 133-minute initial release and a later 163-minute director’s cut with additional material. That longer version is fitting. A story about a fading night scene needs room for what organizations usually cut first: the parts of work that were hard to count, but easy to feel.

When your own industry’s lights change, what proof of value are you still carrying from the old room—and who has decided whether it counts?