Historically, labour abundance has meant innovation scarcity because it gave capital and technology an easier way forward. In the intelligence era of the dawning Liveness machine, that calculation has to change because labour can, in a certain sense, be more than abundant; it can be practically infinite with dampened human downsides.

The change isn’t confined to micro innovation, where one hill-climbs on verifiable tasks. Practically infinite labour also makes macro innovation cheap, because paradigms can be explored in parallel rather than bet on one at a time. What remains scarce is not the jump but the commitment: choosing which of many explored paradigms to grow around.

Even commitment may get cheaper, once paradigms can be simulated and stress-tested before irreversible investment. How far that goes, and what is left for labour, technology and capital to divide between them, may depend on which of Impagliazzo’s five computational worlds we turn out to be living in.