, recently sparked a profound debate by comparing the energy requirements of human development to those of large language models. He argues that the true cost of human intelligence includes twenty years of caloric intake, shelter, and the collective biological history of 100 billion ancestors. This perspective attempts to normalize the staggering electrical demands of
identifies this comparison as a form of data-driven nihilism. When we reduce the value of a person to their energy-efficiency ratio, we strip away the intrinsic purpose of human society. The tech sector's fixation on Return on Investment (ROI) often ignores that financial surplus exists precisely to fund the "inefficient" joys of humanity. We earn capital to spend it on children, relationships, and leisure—none of which offer a traditional market return. This clash of ideologies pits a cold, computational view of the world against a humanist framework.
may match human output at lower marginal energy costs post-training, it lacks the biological imperative and social context that define human intelligence. The risk lies in viewing non-sentient tools as superior simply because they require fewer joules to operate.
Implications for Future Resource Allocation
If the masters of Silicon Valley view humans as inefficient biological processors, fiscal and social policy may shift toward prioritizing digital infrastructure over social safety nets. This transition suggests a frightening future where the metrics of machine performance become the standard for valuing human existence. We must resist the urge to view our species through the narrow lens of thermal efficiency.
