AI won't kill factory jobs, but it will change every manufacturing role: Siemens India CEO

TEXT START: AI will reshape manufacturing jobs rather than eliminate them, with workers needing to continuously upskill as factories become increasingly automated.

The Dissection

This is corporate transition management dressed as labor analysis. Siemens India’s CEO converts an employment threat into a productivity narrative: AI will “enable” workers, workers will continuously upskill, and manufacturing will expand. The article treats new tasks as proof that existing workers remain economically necessary. That does not follow.

The bicycle-to-motorcar and fax-to-internet analogies are structurally weak. Earlier technologies created new industries while leaving substantial human coordination and execution bottlenecks intact. Industrial AI targets those bottlenecks directly: planning, inspection, maintenance, inventory, energy optimization, quality control, and machine operation. The relevant question is not whether some jobs appear. It is whether new jobs appear at sufficient scale, pay, and accessibility to replace the labor made redundant. The article provides no evidence for that.

The Core Fallacy

The central error is conflating productivity growth with employment preservation.

Under the Discontinuity Thesis, AI does not need to eliminate every factory role. It only needs to make human labor economically unnecessary across enough tasks that the mass employment → wage → consumption circuit breaks. A factory can expand output, raise manufacturing’s GDP share, and employ fewer people simultaneously.

The claim that industrial AI must be “100% right” does not protect workers. It creates a stronger case for automation: redundant systems, sensor fusion, formal verification, digital twins, controlled environments, and machine-to-machine supervision can reduce the number of humans required to achieve reliability. Humans become exception handlers, auditors, and liability buffers—servitor roles with shrinking headcount—not the productive center of the system.

The hard framework is straightforward:

• P1: AI becomes cheaper and more capable across cognitive manufacturing work.
• P2: Institutions cannot preserve large-scale human-only production domains against competitive pressure.
• P3: The majority lose access to economically necessary labor, even as output and technical sophistication rise.

Mathur’s “evolutionary process” describes the lag, not the destination.

Hidden Assumptions

• Every displaced worker can continuously upskill at the speed required by advancing systems.
• The new roles will be numerous enough to absorb the old workforce.
• Those roles will remain human-essential rather than becoming the next automation targets.
• Productivity gains will generate enough additional demand to compensate for labor displacement.
• Indian firms can coordinate indefinitely to preserve human-heavy production despite cheaper automated competitors.
• Manufacturing growth measured as GDP share will translate into broad wage income rather than concentrated ownership returns.
• Human oversight will remain necessary instead of being automated, centralized, or reduced to a thin compliance layer.
• “Working alongside AI” means durable productive participation rather than temporary supervision before the next deployment cycle.

None of these assumptions is demonstrated. They are the load-bearing fiction.

Social Function

Primary classification: transition management, with elements of elite self-exoneration and ideological anesthetic.

The message reassures workers while preparing firms to adopt the technology. It frames adaptation as an individual duty—continuous learning—so the institution can later blame those made redundant for failing to keep pace. It also makes AI adoption politically easier by promising that displacement is merely role redesign.

The partial truth is that factories will not become uniformly autonomous overnight. Physical systems, safety regulation, legacy equipment, integration costs, and reliability requirements create real lags. But lags are not reversals. They are hospice care for labor intensity.

The Verdict

This is not evidence that AI will preserve factory employment. It is evidence that Siemens wants AI-driven productivity gains without triggering resistance from the workforce required to absorb the transition.

Manufacturing jobs will change. Some will multiply briefly. The structural direction is still labor compression: fewer people controlling more output, with ownership capturing the gains and workers competing for narrower servitor and exception-handling niches. India’s productivity gap is not a shield. It is the backlog of automatable work.

The CEO is describing the runway to obsolescence and calling it a career path.