AI / Automation and the false promise of efficiency

The tech industry is currently promoting efficiency and cost-cutting as key metrics for Generative AI. As a technocrat, I appreciate the potential of GenAI to streamline bureaucracy. However, we should also consider its original purpose: to comprehensively explore a topic and incorporate diverse perspectives into decision-making.

In an era of increasingly intelligent systems, it might seem counterintuitive to prioritize these seemingly outdated approaches. The reason, however, is that we designed these systems to be human. Bureaucracy, in essence, mirrors our actions and insecurities, and serves a necessary function in society. A just system, I would argue, requires these checks and balances, but we should strive to navigate them with greater ease and enjoyment. To achieve this, we need to focus on who these systems are designed for, rather than solely on what we are building.

The same principle applies to automation. While automation might reduce the number of clicks required, it often conceals underlying complexities.

Each layer we introduce adds more complexity, and I believe we should only add complexity when it enhances the system's ability to be human.

Anything else is not worth our time, especially as the future will increasingly demand that we address these fundamental questions: For whom are we building these systems? Who should benefit from them?