** AI Winter is Coming **
AI is a technological revolution. We take for granted that such revolutions occur. Yet understanding their economic and societal consequences remains elusive. Looking backward is the only reliable compass, and today’s pattern is familiar: a new general-purpose technology appears, spreads unevenly, and begins the long, messy process of reorganizing everything around it. * “When general purpose technologies exist, they are a model that can be followed by all, but [whose] configuration takes time, about a decade or more.” *
The framing here is cyclical. Since the modern era, the sequence resets about every 50 years: technological revolution, “financial bubble–collapse–golden age–political unrest, then renewal.” If we date the present cycle to the mid-1990s, we’re only halfway through tthe timeline—yet the echoes of that decade are loud enough to feel like repetition rather than rhyme.
The early stage is always disorienting. The old order frays, and “exhaustion of a paradigm brings with it both the need for radical entrepreneurship and the idle capital that comes with.” It’s a portrait that fits 2022 a little too neatly. Productivity spikes arrive in lumpy bursts—“The quantum jump in productivity brought about by technological revolution leads during this period”—but progress is uneven, prone to mania.
Finance, predictably, rushes in. “The development and diffusion of each technological revolution tends to stimulate innovations in finance.” Sometimes this accelerates useful deployment; often it distorts it.
The “frenzy phenomenon” becomes “a huge process of income redistribution in favor of those directly or indirectly involved in the casino, which funds the massive process of creative destruction in the economy.” The diagnosis is blunt: “Financial markets are not immoral, they are amoral.”
What actually turns a breakthrough into a true revolution is slower, quieter work. “What’s needed for the revolution to fully blossom are the thousand and one small details, sub-technologies.” These accumulate until production capital—not speculative capital—takes the lead. “Production capital ensures the full deployment and widest spread of each technological revolution.” Only then does the system stabilize into a genuine golden age, and only then do its political contradictions start to surface.
Hobsbawm’s warning lingers at the edges: “it is the power to recognize the turning points that will help economists, politicians, and businessmen prepare for the next war, not the last.” The challenge is timing. Revolutions are obvious in hindsight and noisy in real time. Recognizing inflection points is less about forecasting than noticing when the background noise changes pitch.
If this cycle holds, we’re living in the hinge years. Winter is real and will come, but so is the dream of spring.