The calculator case is cleanest. Competent numeracy before mechanical and electronic calculation required multiplication tables extended to three digits, long division performed in the head, estimation skills built through thousands of hand calculations. These constituted a cognitive palace as real as any orator's memory system. The electronic calculator emptied this palace in a generation. The freed capacity was redirected toward higher-level mathematical reasoning — real gain. The palace, however, was gone. Estimation intuition did not migrate to the calculator. It dissipated.
The GPS case is neurologically measurable. Véronique Bohbot's research documented that habitual GPS users showed reduced hippocampal gray matter compared to people who navigated without electronic assistance. The spatial memory palace was physically shrinking. The brain, which allocates resources according to demand, was reallocating the neural territory spatial navigation had claimed. This is not metaphor. The palace is a physical structure; when it empties, the organ changes.
The search engine case is more radical. Sparrow, Liu, and Wegner's 2011 study documented what they called Google effects on memory: when people expected future access to information through a search engine, they were significantly less likely to encode it in memory. The palace was not demolished. It was never built. The expectation of external storage preempted internal storage at the level of encoding.
The programming case is unfolding now. Segal documents engineers who realized months after adopting Claude Code that their architectural confidence had eroded. They could not explain why — the knowledge was still accessible, still one prompt away. What was gone was the inhabitation. The pattern predicts what follows: the next generation of programmers, learning through AI from the start, will never build the programming palace, will not miss the architectural intuition they never had, and will lack — without knowing they lack it — the specific form of understanding only internalization produces.
The pattern was first articulated in its full generality by Yates's history of the memory tradition and its emptying after Gutenberg. Subsequent research in cognitive science — the calculator research, Bohbot's hippocampus studies, the Sparrow-Liu-Wegner Google effects paper — confirmed the structural regularity across technological generations. What had been narrative observation became measurable neurological phenomenon.
Substitution phase. The new tool can carry information the palace carried — often more accurately, always more permanently.
Atrophy phase. Necessity of the palace diminishes; training motivation collapses; the cognitive muscle decays through disuse within one generation.
Preemption phase. The next generation never builds the palace, because the expectation of external storage preempts internal encoding at its neural root.
Redistribution phase. Capability expands across a wider population as the barrier to entry falls — real democratization accompanied by real devaluation of the internalized expertise.
Neuroanatomical reality. The palace is not metaphor; its emptying involves measurable changes in brain structure — shrinking hippocampi, un-built synaptic architectures.