The AI revolution is not unlike the Industrial Revolution and every other technological and digital upheaval that preceded it. The transatlantic slave trade enlarged the pool of investable funds that produced a trade pattern that fostered industrialization in Europe.

This in turn nurtured plantation societies in the West Indies. Likewise, the political and economic geography of the AI revolution is creating opulence in some places and “shadowlands” or a “Fourth World” of “black holes” of social exclusion in others.

On both the plantation and the platform, risks and costs of production are downloaded onto the worker. While platform labour appears to offer healthy degrees of freedom, it shares with the plantation the instability, precariousness, and temporality of plantation labour.

The shaping of platform labour around “gigs” enables the capitalists to keep wages down. Platform workers face insecure income, labour insecurity, and digital surveillance. AI is humans. People on the margins of society everywhere are picking up “gigs”. The neoliberal outlook of the California Ideology promotes an inclusive form of capitalism, “for the common good”.

However, the pieceworkers in emerging economies become interpellated as a new customer base for the benefit of the Global North. Platform workers engaged in content moderation and data labelling are paid per task. The governance of this type of work happens via protocols. Work tasks are fragmented into tiny bits. Each segment is simple, but the whole leads to feelings of disillusionment and precarity.

In August 2023, the UN estimated that more than 120,000 people were trafficked to work in cyber-slavery camps. They found themselves in walled jungle fortresses with barbed wire fences. At these cyber scam centres in obscure border towns, underworld crime syndicates conduct a range of digital transactions using fake apps to display false investments and profit information.

The trafficked workers build platform accounts using a VPN, AI Apps, and 3D video cameras. The issue of slave labour was woven into Marx’s analysis of ancient and modern social formations and was deeply intertwined with his treatment of wage labour.

In his chapter on “The Genesis of the Industrial Capitalist” in volume one of “Capital,” Marx concluded that “the veiled slavery” of the wage labourers in Europe needed the unreserved slavery of the New World as its plinth.

Marx traced the direct dependence of the British Industrial Revolution on New World slavery in the growth in the number of slave ships. He noted that in 1730, Liverpool employed 15 ships in the slave trade. By 1751, 53; in 1760, 74; and in 1770, 96 vessels were in use. By 1792, the number of ships reached 132.

But despite his clear observations, Marx never wrote a treatise on slavery. Moreover, Marx never understood Saltwater Slavery. His treatment of slave labour systems was wide-ranging, encompassing elements such as ancient Greek and Roman slavery, debt slavery, the rise of slavery as a “second type of colonialism”, and slaveowner capitalism as the basis of a distinctive system of accumulation, including its role in the development of capitalist management.

Marx did not understand the rebellions and slave revolts aboard ships. Driven by despair, thousands destined for “shadowlands” in the West Indies jumped into the Atlantic in chains. Max Weber in his text, “The Protestant Ethic and the Spirit of Capitalism” (1930, p. 123), noted that modern society can force humans into a metaphorical “iron cage” where spontaneity evaporates.

Weber explains that this happens with technological advancement simply because new tools have technical superiority over other forms of organization. And so AI models can potentially promote unhealthy rationalization and commodification. In this sense, AI capitalism is not unlike industrial capitalism. The means and the machines are new. But the consequences converge.

The management and delivery of citizen-facing services by an elite group of technical experts to create public value has a long political history. The idea of a technocracy reverberates with Plato’s belief in philosopher kings who have “true knowledge” that will allow them to govern with political skill, temperament, power, and philosophical knowledge.

It is also a central idea in the work of Henri de Saint-Simone. Saint-Simone’s unique French-style model of economic planning encourages the reduction of friction, precision, speed, unambiguity, discretion, knowledge, continuity, and unity.

In technocratic social systems, experts are given the power to manage and govern. The peril here is that the rapidly evolving possibilities that AI-Assemblages have created can make humans believe that they need certain solutions simply because they are technologically possible. But this can create “false needs.”

Governments are now gradually moving AI from the back office to the front. Bureaucracies are now actively seeking to deliver frictionless citizen-facing services by putting controls in place to mitigate drift and hallucinations.

AI-Assemblages already allow some governments to frame and solve numerous bureaucratic problems, and nurture an orientation to excellence in service delivery grounded in a European Enlightenment-era perspective, that puts a premium on progress by boosting public value.

Behind this constellation of beliefs and practices is a presumed benefit from greater access to data, data mining, and data refining. But technology is never neutral. We are always inside the things we make.