Hours Without A.I (www.nytimes.com)

A.J. Jacobs spent 48 hours deliberately avoiding not just generative AI like ChatGPT, but any product or service that used machine‑learning—whether in production, marketing, or distribution. The experiment forced him to dodge more than obvious apps: he found it surprisingly hard to avoid everyday systems such as electricity, running water and other utilities, because many modern infrastructures and supply chains already embed ML for forecasting, optimization, and automated control. What began as a tech detox quickly became a demonstration of how deeply algorithmic systems are woven into the physical world. For the AI/ML community this is a practical reminder that models aren’t confined to flashy chatbots; they power recommendation engines, logistics, sensor-driven maintenance, dynamic pricing, and smart‑grid operations that shape daily life. Jacobs’ experience underscores two technical and policy implications: the difficulty of tracing where ML is used (provenance and labeling), and the need for auditability, opt-out mechanisms, and resilience engineering for systems that civilians rely on. The episode prompts researchers and regulators to prioritize transparency and human‑in‑the‑loop safeguards as algorithmic automation continues to expand across unseen layers of infrastructure and commerce.