AI Shifts Scarcity From Output to Inputs and Human Moments

Artificial intelligence does not eliminate scarcity but rather relocates it, primarily impacting physical resources such as water and unique human experiences. In regions like Texas, the demand for water by data centers is projected to surge dramatically. By 2030, data centers could consume up to 399 billion gallons of water annually, a significant increase from 49 billion gallons in 2025. This escalating water usage has the potential to lower Lake Mead, the nation's largest reservoir, by over sixteen feet in a single year, highlighting the substantial physical input required to power AI's "effortless output." The cooling mechanisms for these machines are increasingly situated in areas with limited water resources.

Concurrently, as AI-generated output becomes cheaper and more abundant, the value of unique human experiences and moments increases. In 2025, Live Nation reported a record 159 million concert attendees, with revenue exceeding $25 billion. For the first time, more attendees came from outside the United States than from within, indicating a growing global demand for live events. The cost of attending a live event, a distinctly human experience, continues to rise as AI commodifies other forms of output. These two trends—increased demand for physical resources and the rising value of human experiences—are intrinsically linked, representing a fundamental shift in what is considered scarce.

This relocation of scarcity is often overlooked by traditional business metrics. Most companies focus on operational questions regarding cost reduction and output acceleration through AI, which are valid but incomplete. They fail to address the more profound question: when production becomes abundant, what truly becomes scarce, and who retains ownership? The initial impact on the labor market provides a clear illustration of this mechanism. Following the introduction of ChatGPT, demand for specific tasks that AI excels at saw a sharp decline. A study of over three million job postings on a global freelancing platform revealed that within months, demand for translations into Western European languages dropped by approximately 30%. Similarly, demand for commodified writing tasks, such as "About Us" page content, was halved. This shift affected even highly experienced and well-compensated freelancers, demonstrating that AI's impact extends beyond simple cost savings to fundamentally alter labor market dynamics and the perceived value of different skills.