Naming something gives you power over it — and also closes down what you can see. Alfred Korzybski, the Polish-American engineer turned philosopher of language, spent the 1930s arguing that the map is never the territory: the label we attach to a thing is not the thing itself, and every label we use subtly amputates part of what's actually there. This matters more acutely in technology than anywhere else, because the names we give to tools — 'platform', 'AI assistant', 'automation', 'disruption' — are not neutral descriptions. They are pre-decisions. Call something a 'platform' and you immediately inherit a playbook: network effects, two-sided markets, scale. You stop asking whether the underlying dynamics actually fit. The label did the strategic thinking before you arrived. Korzybski's corrective was what he called 'index numbers' — the discipline of treating every instance as unique until proven otherwise. Not 'AI' but 'this specific AI system, doing this specific task, for these specific users, with these specific failure modes.' The abstraction is useful for speed. It costs you precision exactly when precision matters most — which is when the playbook stops working and you can't explain why.

Name the last technology you adopted or recommended — what did the label make you stop asking about?

Drawing from General Semantics — Alfred Korzybski