The Continuity Thesis

In the AI Story

The continuity thesis is deliberately designed to offend common sense. The commonsense view of technology organizes itself around discrete inventions and their identifiable inventors. Watt invented the steam engine. Bell invented the telephone. The Wright brothers invented the airplane. This view is pedagogically convenient, narratively satisfying, and almost entirely wrong. Basalla's evidence, accumulated across hundreds of cases, shows that the 'invention' is always a modification — a recombination of existing elements, a variation on prior artifacts, a node in a chain rather than the start of one.

The thesis has specific implications for how we describe technological change. The word revolution, deployed as though it were a neutral descriptor, is in fact a theory of history disguised as a description. To call something a revolution is to claim that the new has no meaningful connection to the old, that a line has been drawn, that the world on one side of the line operates according to different rules than the world on the other. Basalla's entire project was the demonstration that this claim is almost never true. Every transition that the popular narrative treats as revolutionary, examined with patience, dissolves into a continuous process of accumulation whose cumulative effect appears sudden only from a distance.

The AI case exemplifies the thesis with unusual clarity. The transformer architecture that made GPT and Claude possible was introduced by Vaswani and colleagues at Google in 2017, but the attention mechanism at its core descended from earlier sequence-to-sequence models, which descended from recurrent neural networks, which descended from the backpropagation algorithm rediscovered in the 1980s, which had antecedents in optimization mathematics stretching back decades. Each step involved the recombination of existing techniques. None was itself unprecedented. The novelty lies in the specific configuration, not in the creation of something from nothing.

The thesis produces a specific psychological consequence. When a society understands itself as witnessing an evolution rather than a revolution, the accumulated experience of prior transitions becomes a resource. Institutional wisdom remains relevant. The past illuminates the present rather than being invalidated by it. And the process, being continuous, is amenable to intervention at every stage. This is the practical payoff of the continuity thesis: it converts the AI moment from an incomprehensible rupture into a legible process, and legible processes can be influenced by the institutional choices that constitute the selection environment.

Origin

The thesis emerged from Basalla's sustained engagement with the case material of technological history. Each case he examined — whether the steam engine, the cotton gin, the automobile, or the electric light — displayed the same pattern of continuous descent from antecedent artifacts. The accumulated weight of evidence forced the generalization. The thesis is not a philosophical preference but an empirical claim grounded in two centuries of documented technological development.

Key Ideas

Every artifact has ancestors. There are no immaculate conceptions anywhere in the record, and the apparent exceptions always dissolve under closer inspection into lineages.

The word revolution is a theory, not a description. Calling something a revolution claims discontinuity. The claim is almost never accurate.

Perceptual rupture and processual continuity coexist. The experience of a threshold can be real even when the mechanism producing it is continuous throughout.

Continuity enables intervention. A continuous process is amenable to institutional influence in ways that a genuine revolution would not be.

The thesis is load-bearing. If continuity fails, the rest of Basalla's framework fails with it. If continuity holds, the anti-heroic view of technology follows.

Debates & Critiques

The thesis faces its sharpest challenge when applied to transitions where the pace of change accelerates dramatically — the early industrial revolution, the digital transition of the late twentieth century, and now the AI moment. Critics argue that speed itself can produce qualitative discontinuity even when the underlying mechanism remains continuous. Basalla's response would be that speed alters the experience of the transition and the adequacy of the institutional response, but not the structural fact of continuity. The distinction matters because it determines where the intervention should be directed: not at denying the continuous mechanism but at accelerating the institutional adaptation that the continuous mechanism requires.