Saying Goodbye to AI Anxiety: How Black-Collar Workers Are Forged (Qualities)

Translator’s note. “Black-collar” (黑领), in this author’s series, names a worker whose value in the AI era lies on the boundary between humans and AI systems — designing the limits within which AI is allowed to operate, tending the artificial nature that AI has created, and representing human interests when the system overreaches. The previous essay was the “tools” piece; this one is the “qualities” piece.

About the author. Chun Xia is Founding Partner at TSVC.

Original Chinese version. This essay was first published in Chinese on WeChat. Read the original →.

In the AI era, becoming the kind of person who can make judgments on behalf of human beings.

The previous essay was the tools piece on the black-collar worker. It discussed the worker’s three faces — boundary designer, gardener of the artificial nature, representative of human interests — and the arsenal each face should be equipped with. But there is a more fundamental question: into whose hands should this power be placed? Who is qualified to set the boundaries of AI? Who can press the pause button when the system spins out of control? Who bears the consequences when efficiency and human interest collide? This is not a question of skill. It is a more essential question: what kind of person is fit to be a black-collar worker?

1. The Knowledge Foundation of the Black-Collar Worker

In 2025, a Google DeepMind job posting drew attention: the “AI Governance Researcher” position listed PhDs in philosophy, ethics, law, and political science as preferred qualifications, rather than the traditional computer-science background. This is not a “comeback for the humanities” — it is engineering reaching its own boundary. When code cannot answer the question “what should be done,” knowledge from the humanities and social sciences becomes indispensable.

I have lived through a similar cognitive break myself. By 2006, MySpace, Facebook, and YouTube — these social-media platforms — were already taking off. I found that I could not understand them with an engineer’s logic. These products did not improve efficiency, and they had no clear functional value. Later I realized: the problem was not in the products. It was that my interpretive framework was wrong. So I stopped, withdrew for a year, and concentrated on studying the humanities — beginning with McLuhan’s media theory and reading systematically through different schools of postmodern thought. In hindsight it was simple: social media is not just a tool; it rewrites the relationships among people, and reshapes the relationship between people and technology. Today, in front of AI, I have the same feeling. Engineering thinking alone is not enough. After two decades of cross-disciplinary accumulation, I have gradually built a framework I call “techno-anthropology” — a large-scale, cross-disciplinary mode of analysis. The black-collar worker needs to make this kind of cognitive crossing.

Most of the problems the black-collar worker faces have no formula for their solution. Fairness, justice, rights, interests, the good — these concepts cannot be derived directly from code. They require an entire body of humanistic and social-scientific knowledge to support. The disciplines below provide the basic thinking tools for the black-collar worker. The black-collar worker does not need to be an expert in each one; they need to understand the way each discipline asks its questions, and to translate those questions into rules, boundaries, and audit standards for AI systems.

Philosophy (especially ethics) helps the black-collar worker draw red lines that cannot be crossed, and to fall back on judgment in the edge cases the rules cannot cover. Law provides the framework for defining rights, procedural justice, and the assignment of responsibility — when AI causes harm, is the developer responsible, the deployer, or the user? Political science trains the awareness of checks and balances, and the cross-border thinking required for global governance. Sociology and anthropology allow the black-collar worker to see the way technology embeds itself into the social structure, and how “fairness” carries different meanings across cultures. Psychology reveals the cognitive biases of human beings — why we are easily manipulated by an AI’s “confidence score,” and how to stay clear-headed under high-stakes arbitration. Economics (the institutional branch) teaches the black-collar worker to understand incentive alignment, the internalization of externalities, and the dilemmas of collective action. History offers a long-horizon reference, so that one is not swept up by short-term efficiency. Communication studies trains risk communication — when something goes wrong with a system, how do you explain it to the public and repair trust? Management covers organizational design, change management, and cross-departmental negotiation. Education helps the black-collar worker design training systems, so that the entire team evolves together. Cognitive science offers a framework for the human–machine division of labor: AI is good at pattern recognition but lacks common sense; humans are good at causal reasoning but tire easily. Statistics and data ethics handle data bias, causal inference, and the trade-off between privacy and utility. International law and human-rights law provide a minimum global ethical framework for transnational AI systems.

2. Soft Literacy: Capacities That Cannot Be Acquired from AI or from Books

Disciplinary knowledge arms the cognitive layer. But a black-collar worker who can truly hold their ground also needs a set of non-academic, non-intellectual capacities — things that cannot be acquired from AI or from books, but only from being seasoned in the real world.

The black-collar worker’s work is not in front of a screen. It is in front of people: persuading engineers to revise a model, negotiating budget with executives, explaining a decision to the public, building trust with regulators. A technical genius who is socially clumsy may design a perfect AI-governance framework, and yet no one will execute it — because they have not earned the trust. Deep listening, nonviolent communication, the building of cross-hierarchy networks — these capacities are often harder to acquire than technical knowledge, and easier to overlook.

AI governance will, sooner or later, enter public discussion. In ordinary times, the black-collar worker needs to be a thought leader, producing content steadily and building professional reputation. At critical moments, they need to know how to handle a public-relations crisis — when something fails, neither passing the buck nor evading, but explaining the problem and the remedy clearly. Content production, understanding the propagation logic of each platform, awareness of crisis communication — these are no longer avoidable parts of the black-collar worker’s toolbox.

3. Public governance and political participation

The rules of AI will, in the end, be embodied in laws, executive orders, and industry standards. The black-collar worker cannot be merely a passive rule-follower; they should be a co-author of the rules. The founders of San Francisco’s “algorithmic transparency” advocacy organization offer one prototype: they are not politicians, but they can mobilize citizens, debate with the city council, and become the first generation of “public black-collar workers.” Understanding how government works, submitting high-quality feedback in policy consultations, organizing community discussions — these capacities determine whether the black-collar worker can move from inside the corporation to the broader space of governance.

4. Artistic interest and aesthetic sensibility

Aesthetics is not decoration. It is another foundation of judgment. A person without aesthetic sense will struggle to judge whether the output of an AI system is “appropriate” — whether it is the visual oddness of a generated image or the emotional dissonance of an algorithmic narrative. More importantly, art teaches a person to tolerate ambiguity, contradiction, and the unfinished — which is precisely the real-world condition the black-collar worker faces every day. The final decision is rarely a choice between black and white; it is a search for an acceptable balance in the gray.

5. Sports and physical literacy

The work of a black-collar worker is intense: long stretches of focus, meetings across time zones, crisis management. Without the body’s stamina, all of it collapses to zero. But sports offer far more than fitness. Endurance sports train one to remain steady under sustained pressure. Competitive sports train calm decision-making in conflict. Team sports train trust and collaboration. More importantly, sports teach a person to accept failure, to respect rules, and to push through one more time at the limit — none of which can be acquired from reading. They can only be internalized through bodily failure and bodily recovery.

6. The practical wisdom of daily life

Some seemingly small life domains provide unexpected mental training for the black-collar worker. Cooking and housework let one understand end-to-end management, “from raw material to finished product.” Cross-cultural travel trains navigation in chaos, building connections with strangers, and a curious stance toward difference. Volunteer service offers training in empathy and in acting without expectation of return. Gardening and farming let one understand slow growth, the uncontrollable, and seasonal rhythm — which is precisely the mental disposition the gardener needs when facing the artificial nature. Crafts and repair train patience and sensitivity to materials. Chess, board games, and esports train strategic thinking and decision-making under uncertainty. Meditation and mindfulness help one stay calm in high-pressure arbitration. These experiences may seem unrelated to AI, but at the critical moment, they determine the quality of one’s judgment.

7. Emotional and psychological resilience

The black-collar worker stands on the boundary between humans and AI, and is destined to be unwelcome on both sides. The engineer thinks you are getting in the way. The business side thinks you are too cautious. The public thinks you are the company’s spokesperson. This position requires the courage to be disliked, the capacity to disassemble emotion (recognizing the fear underneath the anger), and a support system (peer circles, mentor groups). More importantly, it requires recoverability — the ability to debrief and start again quickly after failure. Without these psychological resources, it is hard to last in this kind of long-term pressure.

3. Judgment at the Critical Point

Equipped with knowledge and soft literacy, the value of the black-collar worker ultimately shows itself at the critical point.

Take autonomous driving. A car is moving on a rainy night when a pedestrian crosses the road, while another car appears on the left. The system has to make a choice in tenths of a second. The engineer can preset rules, but cannot exhaust every situation. In the end, someone must decide on the system’s order of priorities. This is not a technical question. It is a value judgment — and that is precisely the territory of ethics.

When medical AI mistakes a cancer patient for an ordinary inflammation, the attending physician faces a choice: trust the AI for the sake of efficiency, or override it and bear the responsibility personally. The black-collar worker must answer in advance: under what conditions must a human override the AI’s judgment? This question touches on legal questions of liability and on cognitive-science questions about the human–machine division of labor.

A company uses AI to optimize its cost structure, and the system recommends laying off a group of “lowest-efficiency” employees — but those people happen to be the company’s most experienced middle managers. The system does not automatically recognize this kind of structural risk. The black-collar worker must see it from the perspective of sociology and history: the social consequences of technological transformation often lie outside the optimization function.

4. Getting the Right Judgment Executed

The black-collar worker is not necessarily the smartest person. But they must be the person who can make a correct judgment actually be carried out. Suppose an AI team decides to launch a new system, and you have identified a risk inside it. You need to persuade the engineers, the product managers, the executives, and sometimes even the public. Otherwise the system goes live as planned. I have seen many people with the right judgment watch the wrong outcome unfold simply because they could not communicate effectively. This means the black-collar worker needs three kinds of language: technical language to persuade engineers, commercial language to persuade executives, and public language to communicate with society. It is a cross-domain translation capacity.

5. The Capacity for Anti-Efficiency

This may be the most counterintuitive point of all. The work of a white-collar worker is to improve efficiency. One of the jobs of a black-collar worker is to keep efficiency from overstepping its bounds. For example: a company plans to fully automate its customer service. Technically, this is entirely feasible. But a black-collar worker may argue for keeping a portion of human agents — not for reasons of efficiency, but because emotional issues cannot be handled by automation, edge cases cannot be covered, and brand trust may be damaged. This is a deliberate decision to lower efficiency. In an organization that worships efficiency, this kind of decision requires real grounding and a clear stance.

6. The Black-Collar Worker, Too, Must Be Constrained

If unconstrained, the black-collar worker can become a new center of power — possibly more dangerous than the white-collar worker, because they decide the boundaries. So a system of checks and balances must be built around them. The decision-making process must be auditable, so that when something goes wrong it can be traced back. The black-collar worker themselves must be accountable; power without accountability eventually corrupts. And there must be multi-party oversight — engineers, users, third-party institutions, and regulators all participating, forming a network of mutual constraint. The black-collar worker is not the end point. They are one component in a new institutional architecture.

Conclusion

In Sequoia National Park, the redwoods grow to over a hundred meters and live thousands of years, but their roots reach less than two meters deep. They stand because they grow in groups: their roots intertwine underground and support each other. The black-collar worker is the same. They are not lone heroes. They are nodes in a network of mutual constraint and mutual trust.

The real black-collar worker is not chasing personal “rightness.” They are participating in the construction of an institutional framework in which “rightness” emerges as a system property. A black-collar worker without humanistic literacy is just another form of white-collar worker — still a passive instrument, only the object of operation has changed from Excel to an AI-governance dashboard. A complete black-collar worker is a person who has read, has seen, has practiced, has loved, and has hurt.

The white-collar worker is a product of the institutional age. The black-collar worker is a product of the AI age. In the past, white-collar workers kept the world running. The black-collar worker will decide how the world runs. The white-collar worker solves problems. The black-collar worker decides which problems are worth solving.

The forging of a black-collar worker is, at root, a course of life — not just a course catalog.