The Planned Obsolescence of Capabilities: AI, Health, and the New Frontier of Inequality
By: MSc Victor Piriz Correa, MD, MPH. Lorena Leon Directora de Salud Ocupacional de SICs
Planned obsolescence is typically associated with the deliberate design of a product's end-of-life. For years, we conceived it primarily in relation to household appliances, mobile phones, or computers: objects that once lasted decades and today seem designed for replacement in increasingly brief cycles. However, limiting this concept to the goods industry is insufficient. A closer analysis compels us to recognize various forms of obsolescence: natural, programmed, perceived, planned, and even decreed. Perhaps the most perilous in the digital age is the least visible: perceived obsolescence, which arises not because something ceases to function, but because the system decides it is no longer of value.
This logic, once reserved for consumer markets, is now being projected onto human labor. Our concern is that generative artificial intelligence could become an engine of obsolescence not only for products but for professional skills, career trajectories, and entire economic structures, before workers and nations have sufficient time to adapt. We are entering a stage where the risk is no longer merely that a device becomes obsolete, but that a worker is perceived as obsolete for failing to keep pace with technological updates.
This issue is particularly grave in the healthcare sector. There, AI must be envisioned not as a replacement, but as an augmentation. This distinction is decisive. If used as a complement, it can reduce administrative burdens, mitigate burnout, and enhance clinical capabilities. If used as a validation criterion for the professional, we enter a logic of human disposability: the value of the clinician is measured by their familiarity with the latest interface or algorithm, rather than by their clinical judgment and their bond with patients. This is the planned obsolescence of capabilities.
The environmental consequences of classical planned obsolescence are well-documented. The surge in electronic waste (e-waste) reached 62 million tonnes in 2022 [1]. This is not merely an inconvenience; we are discussing heavy metals and toxic compounds that degrade soil and water, affecting human and ecosystem health [1,2]. Therefore, planned obsolescence is no longer just a market distortion but a failure of global governance.
The novelty of our time is that this logic is shifting from objects to people. The ILO’s Working Paper 166 (2026) illustrates a critical asymmetry: in many developing countries, workers vulnerable to automation already have sufficient connectivity to feel the impact rapidly, while those who could benefit from AI productivity face digital infrastructure bottlenecks [3]. This is the infrastructure trap.
In healthcare, this threat acquires an even more delicate ethical dimension. The sector works with trust, vulnerability, and moral responsibility. Recent literature insists that AI can be a useful adjunct, but not a substitute for clinical reasoning [5]. When a system values algorithmic speed over clinical judgment, a form of silent institutional violence emerges: the professional is treated as an interchangeable part.
Furthermore, AI is not without material cost. It requires data centers, energy, water, and critical minerals [6]. If this footprint is part of an economy of accelerated turnover, AI becomes a multiplier of material obsolescence. Our position must be clear: AI in health must be an augmentation, not a replacement. We need governance capable of preventing technological cadence from dictating human validity. We need social dialogue, infrastructure investment, and a new ethical framework that recognizes that neither professionals nor countries are expendable hardware.