Seventy-to-One

In the AI Story

The number is the empirical instantiation of the pattern Deaton's career has documented: genuine progress in the aggregate, widening gaps in the distribution. The aggregate improvements in human welfare since the industrial revolution are real and substantial. Life expectancy in wealthy nations approximately doubled. Extreme poverty fell from encompassing the vast majority of the human population to levels that would have seemed utopian to any eighteenth-century observer. Infant mortality collapsed. Literacy spread. These improvements reached hundreds of millions of people in the poorest nations as well, raising their absolute conditions beyond anything their ancestors experienced.

But the rate of improvement differed dramatically across nations. The populations that escaped first — in Britain, then Western Europe, then North America, then pockets of East Asia — accumulated advantages that compounded over generations. Better health produced more productive workers, who generated more surplus, which funded further improvements in health, education, and infrastructure. Better institutions attracted investment and talent, which strengthened the institutions. The early escapees pulled ahead at rates the late arrivals could not match, producing a divergence that widened even as the late arrivals made absolute progress.

The Segal foreword to this volume frames seventy-to-one as the number that 'broke his confidence' when he encountered it. The book he had written, The Orange Pill, had celebrated the democratization of capability that AI tools enable. Seventy-to-one forced the question the Orange Pill book had not foregrounded: what happens when early beneficiaries of a new technology compound their advantage so quickly that later beneficiaries can never catch up? The pattern, Deaton's career demonstrates, is the default. Departure from the pattern requires deliberate institutional construction.

For AI, the stakes of the pattern are measured in the specific currencies Deaton has spent his career counting: years of life expectancy, children's educational attainment, the capacity of individuals and communities to participate in the economic and social life of their societies. If the AI transition produces a ratio analogous to seventy-to-one — compressed into decades rather than centuries because the technology moves faster — the resulting divergence will shape the global distribution of capability for generations. The dams must be built while the transition is underway, not after.

Origin

The seventy-to-one ratio is documented in Deaton's The Great Escape (2013) drawing on the historical income data compiled by Angus Maddison. The five-to-one pre-industrial baseline draws on the same source. The number has become a touchstone in contemporary discussions of global inequality.

Key Ideas

The aggregate conceals the distribution. Progress can be real at the species level while specific populations fall further behind.

The gap widened during the escape, not because of stagnation but because of divergent rates of improvement.

Early advantages compound. The populations that escape first accumulate resources, institutions, and capabilities that make subsequent escape easier for them and harder for others.

The pattern is the default. Absent deliberate institutional construction, technological transitions produce widening rather than narrowing gaps.

AI accelerates the mechanism. The compressed timescale of the AI transition means the default pattern operates faster than institutional responses can match.