Installation and Deployment Phases

In The You On AI Field Guide

Perez's framework emerged from detailed historical study of five technological revolutions since 1771: the industrial revolution, the age of steam and railways, the age of steel and electricity, the age of oil and mass production, and the age of information and telecommunications. In each case, Perez identified the same two-phase pattern, separated by a turning point typically associated with financial crisis.

The installation phase is characterized by specific features: financial capital dominates, speculative bubbles form and burst, infrastructure is built ahead of demand, and the economy bifurcates between the new sector (growing rapidly) and the old economy (stagnating). The deployment phase replaces these features with their complements: production capital dominates, broad-based prosperity becomes possible, institutional arrangements channel the technology toward mass adoption, and the economy integrates the new technology across all sectors.

The turning point between phases is, in Perez's framework, typically associated with financial crisis. The 1930s depression was the turning point for the age of steel and electricity. The 2008 financial crisis was, in Perez's reading, the turning point for the age of information. The crisis exposes the limits of the installation-phase arrangements and creates political pressure for the institutional innovation the deployment phase requires.

The framework's application to AI is contested. Some readings — including Andreessen's — treat AI as a continuation of the information age, meaning the current moment represents deployment-phase dynamics playing out. Other readings — including Perez's own recent work — treat AI as potentially the opening of a sixth technological revolution, meaning the current moment represents a new installation phase. The distinction matters for policy: deployment-phase institutional innovation looks very different from installation-phase experimentation.

For the Andreessen — On AI volume, the framework provides analytical distance from the immediate intensity of the AI discourse. The Software Death Cross, the speculative valuations, the institutional lag — these are features the framework predicts and specifies. The volume argues that navigating them requires the specific form of institutional innovation Perez's framework identifies, and that the absence of adequate institutional innovation is the primary risk of the current moment.

Origin

Perez developed the framework across a career that began with her 1983 article on technological change and economic cycles, and culminated in her 2002 book. The framework extended and systematized observations by Schumpeter, Kondratiev, and Chris Freeman, organized around the specific finding that financial and production capital play systematically different roles across the phases of each revolution.

Key Ideas

Two-phase structure. Each technological revolution unfolds in an installation phase (speculative, financial-capital-driven) followed by a deployment phase (institutional, production-capital-driven), separated by a crisis-driven turning point.

Financial vs. production capital. The two phases are distinguished by which form of capital dominates — financial capital's preference for liquidity and speculation during installation, production capital's preference for patient investment during deployment.

Institutional innovation requirement. The transition between phases requires the construction of new institutional arrangements — labor law, regulatory frameworks, distributive mechanisms — that the installation phase does not produce on its own.

Crisis as turning point. The passage from installation to deployment is typically triggered by financial crisis, which exposes the inadequacy of installation-phase arrangements and creates political pressure for reform.

AI transition ambiguity. Whether AI represents a new installation phase or the deployment phase of the information revolution determines what institutional innovation the moment requires — an analytical question with major policy consequences.