CONCEPT
Material Substrate of Computation
The physical infrastructure—electricity grids, cooling systems, semiconductor fabs, water supplies—that computation depends upon, invisible to users but thermodynamically non-negotiable.
The material substrate of computation names the complete physical foundation underlying digital information processing. Every token generated by a large language model requires electricity drawn from power plants, silicon fabricated in billion-dollar facilities, water evaporated through cooling towers, and rare earth elements extracted from geographically concentrated mines. Smil's framework insists this substrate is not background scenery but the binding constraint—more fundamental than software capability because without it, the software has nothing to run on. The substrate's visibility to end users is inversely proportional to its importance: the cleaner the interface, the heavier the hidden foundation.
In The You On AI Field Guide
The contemporary AI discourse operates almost exclusively at the level of capability—what models can do, how quickly they improve, which benchmarks they surpass. This focus obscures the thermodynamic reality that every computational operation is simultaneously an energy transformation governed by physical law. A GPU performing matrix multiplication draws electrical current, converts a portion to useful computation, and dissipates the remainder as heat at rates determined by chip architecture