Those who understand AI better are more afraid of losing their jobs - TU Darmstadt

TEXT START: Whether employees are worried about losing their jobs to AI depends largely on their knowledge of artificial intelligence – and on the extent to which their work can be digitalised.

The Dissection

The article is documenting the collapse of the reassurance narrative. Its central finding is that informed workers are not afraid because they misunderstand AI; they are afraid because they understand its reach. Digital-output workers—lawyers, developers, analysts, writers—can see that the product of their labor is becoming reproducible by machines.

The text also converts a systemic rupture into a management problem. It acknowledges that displaced workers lose more than income, including status, structure, and social participation, but directs the response toward training, honest communication, and policymaking. That is an accurate description of the symptoms paired with an inadequate treatment plan.

The survey is meaningful evidence about perception and exposure. It is not, by itself, proof that AI has already achieved permanent cost and performance dominance. But it decisively weakens the comforting claim that public fear is merely an information deficit.

The Core Fallacy

The main conceptual error is treating training as a plausible large-scale answer to the destruction of labor demand.

Training can help a minority move into scarce, complementary roles. It cannot manufacture enough economically necessary work once AI can produce digital outputs at lower cost and greater scale. If the machine performs the valuable function, teaching more people to perform the superseded function only creates a larger queue for declining employment.

The article also preserves the conventional assumption that new work will compensate for eliminated work. Its own wording admits that replacement jobs will probably not appear on anything close to the scale of the losses. That is not a labor-market adjustment. It is the beginning of productive participation collapse.

Hidden Assumptions

• That workers can be retrained faster than firms can automate their tasks.
• That newly trained workers will find sufficient demand rather than compete for a shrinking number of human-complementary roles.
• That digital and physical work are separate permanent categories. Physical labor has a lagging moat, not immunity.
• That companies and policymakers can coordinate stable human-only economic domains at scale.
• That preserving consumption or income transfers would preserve meaningful productive participation.
• That social participation can be restored through employment policy after employment ceases to be economically necessary.
• That fear of job loss is primarily a psychological issue requiring reassurance or explanation, rather than a rational response to declining bargaining power.
• That the transition can be managed without a decisive transfer of ownership and control of AI capital.

Social Function

Primary classification: partial truth functioning as transition management, with a secondary role as ideological anesthetic.

It is not simple copium. The article openly states that knowledgeable people see the danger and that job creation will likely fall far short of job destruction. That is unusually close to the structural truth.

The anesthetic enters through the proposed remedy. “Training and honest answers” makes a system-level displacement process appear administratively manageable. It shifts attention from who owns the productive machines to whether workers have received enough instruction. The language is humane, but the implied bargain is brutal: workers may be told the truth about their dispossession while being offered credentials for occupations that the same machinery is eroding.

The Verdict

This is a useful warning wrapped around an insufficient prescription. Its strongest finding is that the people closest to AI can already see the guillotine. Its blind spot is that awareness, training, and honest communication do not restore labor’s economic necessity.

Under the Discontinuity Thesis, the article is an early autopsy of the wage-to-consumption circuit. It records the loss of confidence before the system formally loses the ability to provide mass productive participation. The workers are not irrationally afraid. They are recognizing that the machine is becoming the employee, while human beings are being reassigned to the waiting room.