Pister, Kahn, and Boser's 1997 DARPA proposal, a complete sensor in one cubic millimeter, was funded in 1998; the RAND 1992 workshop and mid-90s ISAT studies frame it as military from birth. The concept's genealogy is stranger: Stanislaw Lem described the idea in fiction, 1964 and 1985, before any lab. Real-world motes now monitor vineyards, oil tankers, parking; Streetline deployed 12,000 in San Francisco. The 2022 University of Washington Nature paper: programmable, battery-free sensors that travel a kilometer on the wind. The name entered Gartner's hype cycle as its most speculative entrant and stayed real.
The theory: smart dust in the blood, the air, the vaccine, invisible networks already deployed. The documented version is both more boring and more interesting: the technology is real but bulky, power and antenna are the constraints, and the genuine deployments are agricultural and municipal, in the open. The privacy question is legitimate and unresolved: sensors that are hard to detect and harder to remove, by the field's own description, do not ask consent.
The claim: smart dust is the end-state of the Panopticon, surveillance that no longer needs a tower because the dust is the tower, the watched world becoming coextensive with the world; Bentham's inspection principle finally leaving the building and entering the weather. Verdict: the military origin is documented, the dispersal is documented, the trajectory is the obvious extension. The archive's rule from Carbyne applies: the infrastructure is real, the intent is the contested part. Check the forecast. It may be checking back.
If the dust can ride the wind, what can it ride next?