In the Age of AI and Automation, Some Jobs Are Set to Shrink by 2035 - The HR Digest

TEXT START: Which are the jobs that are most likely to shrink in the next decade?

The Dissection

The article identifies the first visible casualties of automation: cashiers, office clerks, customer service representatives, secretaries and administrative assistants, and bookkeeping, accounting, and auditing clerks. It also cites a projected 4% decline in administrative-support employment and a possible 200,600-job reduction among cashiers.

What the text is really doing is converting systemic displacement into a manageable career-planning problem: workers should reskill, employers should supervise AI, and new technology sectors should absorb the displaced. It describes the bleeding without naming the organism killing the patient.

The Core Fallacy

The article treats declining occupations as isolated categories rather than early targets in a general automation cascade. Under the Discontinuity Thesis, AI does not merely eliminate repetitive tasks. Once it becomes cheaper and more capable across cognitive work, it erodes the coordination layers built around those tasks.

Its “human and technology developing in parallel” solution assumes augmentation remains economically superior to substitution. That is not a moral or managerial decision. Competitive pressure selects the cheaper, faster, scalable system. The article’s own feedback loop—fewer workers, weaker talent pipelines, and stronger incentives to automate—points toward acceleration, not confusion.

The 2035 projections are therefore social-lag estimates, not mechanical limits. They measure what institutions currently expect, while P1, P2, and P3 describe what happens when AI superiority, coordination failure, and mass loss of productive necessity converge.

Hidden Assumptions

• BLS projections will remain stable despite accelerating capability and adoption.
• Displaced workers can retrain quickly enough and find comparable demand.
• Growth in healthcare, energy, utilities, and technology will absorb labor at sufficient scale.
• Human supervision will remain necessary rather than itself becoming automated.
• Niche expertise will retain value even as AI makes expertise reproducible.
• The wage-to-consumption circuit can survive after most people lose economically necessary work.
• Employers can freely choose a humane labor model without being disciplined by competition.

Social Function

Classification: partial truth, transition management, and ideological anesthetic.

The article is useful as an inventory of exposed roles, but its “not time to panic” framing softens the mechanism. “Upskilling” transfers responsibility for structural displacement onto workers, while the proposed human-AI partnership disguises the fact that firms automate because labor is a cost and AI capital is controllable. It prepares people psychologically for contraction while preserving the fiction that ordinary career mobility remains a reliable escape.

The Verdict

The article correctly identifies administrative and transaction-heavy jobs as early casualties. Its prognosis is wrong. These roles are not merely “winding down”; they are the exposed edge of P1. A 4% decline is not proof of stability. It is social inertia rationing the amputation.

As AI expands from cash handling and clerical processing into customer service, accounting, scheduling, analysis, and managerial coordination, P2 prevents durable human-only economic domains from being preserved at scale. P3 follows: the majority lose access to economically necessary labor.

“Upskill” remains viable mainly for future Sovereigns who control AI capital or Servitors who become genuinely indispensable to them. For everyone else, the article offers a ladder leaning against a collapsing wall. It is an accurate symptom report and a false systemic prognosis.