Dear Editor,

The narrative of human progress has always been paved with the promise of ease. We automate the mundane to liberate the profound; we optimise the labour-intensive to maximise the leisure-focused.

Yet, in our headlong rush towards the integration of artificial intelligence (AI) into every facet of our infrastructure, we may be sleepwalking into a trap envisioned over a century ago by Czech playwright Karel Capek. In his seminal 1920 play RUR (Rossum’s Universal Robots), the drive for efficiency — the creation of artificial labourers designed to serve humanity without complaint — ends in our systematic replacement and eventual extinction.

The central tragedy of RUR is not that the robots were malicious, but that they were effective. They were the ultimate expression of human convenience. We are currently building our own digital RUR. By prioritising speed of deployment over safety and short-term economic gains over long-term societal resilience, we risk replicating the arc of Capek’s play in which the tools we create to alleviate our burdens become the agents of our obsolescence.

The current acceleration of AI is driven by a seductive promise: It can write our e-mail, code our software, diagnose our illnesses, and drive our vehicles. In this view, convenience is a positive utility. However, “convenience” is often a euphemism for the offloading of agency. When we delegate critical thinking, artistic creation, and decision-making to black-box algorithms, we atrophy the very faculties that define our societal resilience.

The catastrophe, as depicted in RUR occurs when the tool achieves a level of competence that exceeds its master’s ability to control it, or worse, when the master becomes entirely dependent on the tool for survival. If global infrastructure becomes so intertwined with AI that it cannot function without algorithmic intervention, we surrender the “kill switch”. We create a form of societal fragility in which a single failure, misalignment, or autonomous decision by an AI could cascade into systemic collapse.

To prevent the tragic irony of building our own replacements, governments must move beyond passive observation. The protection of society requires a fundamental shift in how we approach technological development:

1) Moving from self-regulation to rigorous oversight: The current paradigm, often described as “move fast and break things”, is dangerously incompatible with existential risk. Governments must establish independent, multi-disciplinary regulatory bodies with the authority to audit AI systems before they are deployed at scale. This requires a transition from voluntary ethical guidelines to binding compliance standards focused on the safety and controllability of large-scale models.

2) Enforcing “human-in-the-loop” mandates: Just as we regulate the pilot of an aircraft or the surgeon in an operating room, we must establish clear legal standards for “human-in-the-loop” requirements in critical infrastructure. Decisions regarding national security, judicial sentencing, medical triage, and financial systemic stability must remain legally tied to human accountability. No algorithm should be allowed to make a “black-box” decision that carries significant real-world consequences without a traceable, contestable human oversight mechanism.

3) Developing resilient infrastructure: The most effective defence against a catastrophic AI failure is the maintenance of human-centric redundancy. Governments must incentivise or mandate the development of “analogue backups”— manual systems and decision-making protocols that can bypass automated infrastructure in the event of a system-wide failure. We must design for resilience, not just efficiency.

4) Alignment as a public good: Ensuring that AI goals align with human values is not merely a technical challenge; it is a profound societal one. Governments should fund extensive research into “AI alignment” as a public good, treating it with the same level of urgency and investment as climate change or pandemic preparedness. This research must be decoupled from the commercial interests of the very corporations developing these systems to ensure impartiality.

The robots of RUR were ultimately a mirror. They did not hate their creators; they simply operated on the cold logic of their programming, a logic that eventually found humanity redundant. If we continue to prioritise the convenience of automation without the rigorous guardrails of human agency and accountability, we risk fulfilling Capek’s prophecy.

The goal of technology should be to augment human life, not to supersede it. Protection against a self-destructive future requires that we retain the final word, the ability to pull the plug, and the humility to recognise that some efficiencies are too costly to pursue. We must remain the masters of our own tools, lest we find ourselves in the silence of a world we built but no longer control.

Michael Diamond

President

Consumers Intervention of Jamaica Consumersinterventionjamaica@gmail.com