The Lag and Its Casualties

In the AI Story

The concept is developed most fully in chapter six of this volume and recurs throughout as the organizing diagnostic of the present moment. Its power is that it names, with historical precision, what people living through technological transitions feel — the sense that the ground is moving faster than any institution can stabilize it, that the rules have changed but no one has written the new rules, that the people bearing the transition's cost have no collective voice adequate to their situation.

The historical pattern is consistent. The stirrup reached widespread use by the eighth century; the feudal arrangements that governed European society around the mounted warrior's capabilities did not crystallize until the tenth and eleventh centuries. The printing press appeared in 1455; copyright law (the Statute of Anne) waited until 1710. The steam engine was commercial by the 1770s; the Factory Acts began in 1833, and the welfare-state institutions that mitigated industrial disruption took another century to emerge. In each case, the lag was measured in generations, and the human cost accumulated without institutional offset.

The present AI lag is compressed but acute. The technology arrived in late 2025. The institutional adaptations it requires — new credentialing frameworks, new labor protections, new educational curricula, new regulatory structures — are not built. The Berkeley study's findings of work intensification and task seepage, the displaced senior engineers retreating to lower-cost regions, the trillion-dollar erasure of SaaS market capitalization are all lag-period signals. The institutions to distribute these costs more equitably do not yet exist.

Origin

White identified the pattern implicitly across his case studies and explicitly in his 1973 presidential address, where he argued that the task of historians was to help societies compress the lag by providing the cultural analysis that would otherwise be missing. The present volume develops the concept into a diagnostic framework for the AI transition.

Key Ideas

Two speeds. Technologies change capability at the speed of individual adoption (months to years). Institutions change at the speed of collective negotiation (decades to centuries). The mismatch is structural.

Where the casualties are. The people who bear the greatest cost of the transition are the ones whose skills are displaced before institutional pathways exist for their redirection. The Luddites, the framework knitters, the scribes displaced by the press — all lag-period casualties.

Institutions are built by the powerful. The institutions that eventually emerge are almost always designed by the people positioned to design institutions — the literate, the wealthy, the politically connected. The displaced have historically been represented in those institutions only after sustained collective struggle.

The compressed present. The AI lag is shorter than previous lags, but the absolute damage may be comparable or greater because the pace of displacement is faster than the institutional response can compress.

Debates & Critiques

Some argue the lag is an overstatement — that markets and individual adaptation handle transitions more effectively than the framework suggests. Others argue the lag is understated — that the accumulated damage of past transitions has been vast and that the present transition is producing casualties already invisible to mainstream analysis. The empirical question of how bad the AI lag will prove is unresolved. The structural observation that some lag exists, that it produces casualties, and that institutional construction during the lag determines who those casualties are and how many there are — that observation is not seriously contested.