US Workforce In 2035: A Few More Workers, A Lot More Output - Josh Bersin

TEXT START: Despite the talk about job losses from AI, we now know the opposite is true.

The Dissection

The article converts aggregate labor projections into a rebuttal of AI displacement. It shifts from jobs to tasks, labels surviving work “Superworker” work, and assumes displaced employees will migrate smoothly into reinvented roles. Its real function is to reassure CEOs and HR departments while marketing a management agenda, a book, and Galileo.

The Core Fallacy

It confuses output growth and aggregate labor shortages with continued mass productive participation. The article’s own figures—3.5% workforce growth against 22% GDP growth—describe rising output per worker. That is not evidence against the Discontinuity Thesis; it is the mechanism of labor decoupling.

A projected shortage of nurses, caregivers, electricians, or plumbers can coexist with displaced clerks, programmers, support workers, and administrators. Labor markets do not automatically transfer people across skills, locations, credentials, wages, or age. “Human-centric” is not a permanent moat. AI can cheapen cognitive support work, while robotics attacks physical work later. The article admits this, then pretends the admission does not undermine its conclusion.

Its claim that displaced workers “won’t be unemployed” is asserted, not demonstrated. Wage premiums for AI-enabled workers may reflect scarcity and ownership leverage, not broad-based worker empowerment. More output requiring fewer workers weakens labor’s bargaining position unless ownership of the productive systems is redistributed—which the article never addresses.

Hidden Assumptions

• BLS projections remain reliable despite accelerating AI and robotics capability.
• Occupation-level growth implies viable transitions for incumbent workers.
• Physical contact and human empathy remain economically scarce indefinitely.
• Automation removes tasks without reducing the number of workers required.
• Rising wages in selected occupations are accessible to displaced workers.
• Management, sales, and teaching growth represents durable human necessity rather than temporary institutional lag.
• Aggregate labor shortages translate into worker power rather than localized bottlenecks and wage stratification.
• Productivity gains flow to workers instead of AI-capital owners.
• The current 830-title taxonomy can capture a nonlinear technological discontinuity.

Social Function

Primary classification: transition management disguised as copium, with prestige signaling and elite self-exoneration.

The article acknowledges genuine lag defenses—aging, demographic decline, healthcare demand, and the persistence of physical work. But it packages temporary servitor niches as proof that the system remains healthy. “Superworker” is the anesthetic: it renames intensified dependence on AI as empowerment and turns structural displacement into an HR training problem.

The Verdict

This is not evidence against the Discontinuity Thesis. It is a lag-phase memo describing how the old system can produce more GDP while needing barely more labor. The headline accidentally states the central danger: a few more workers, a lot more output means the employment-to-consumption circuit is losing its centrality. Human-touch occupations are temporary defenses, not salvation. The durable winners are the Sovereigns controlling AI, energy, logistics, maintenance, and capital; everyone else is being sorted into increasingly replaceable servitor niches.