The AI investment boom – whether or not it is, as I tend to believe, a bubble – is a vivid demonstration of how misleading it is to characterize capitalism as a ‘market economy’. The system’s ideologues constantly prattle on about markets. In their telling, markets transmit information about people’s wants and desires to distant producers, who then orient themselves accordingly. The cult of markets, then, comes down to a claim about the relationship between information and investment. For market enthusiasts, even the best-willed central planner in the world cannot respond adequately to people’s needs, because planning destroys the price system that conveys the information which would allow rational planning to take place.

Leaving aside complex debates about calculation and planning, consider for a moment whether markets play the role they are supposed to do under capitalism. A moment’s reflection reveals that if capitalism could indeed be described as a market economy tout court it would have none of the revolutionary characteristics that it evidently has possessed in the past. The most obvious feature of the system is its tendency constantly to transform the techniques and organization of production, with the purpose of increasing profits in a competitive environment. It is not at all obvious that AI and the associated data centre build-out will lead to such an umwälzung. It is however evident that some important actors think it will, or would at least like to give the impression that they do.

Whatever the ultimate outcome of the current investment boom, one thing is clear. It cannot possibly be a response to information about the existing preferences of consumers. Precisely because generative AI is a radically new technology, the use of which is still being worked out, the investments and the build-out they are fuelling are independent of demand, or at best are based on guesses, which cannot be based on prices. This non-responsiveness to market signals is precisely what makes these investments potentially revolutionary.

Consider the situation in November 2022, just before the release of ChatGPT. The demand for generative AI at that point was basically non-existent. But following the release, huge quantities of investment flowed into developing the technology. Why did all this investment cascade into a business for whose products there was no evident demand? It occurred because the rhythm of capitalist investment is determined by future expectations of profitability, which is largely disconnected from the existing wants and desires of the population, let alone any broader consideration of social needs.

The system of supply and demand in fact plays a subordinate role in capitalism. The most characteristic feature of capitalism is how independent investment is from demand – how much it is geared toward a speculative future, and how little it relies on information. By contrast, a pure market society would be totally unlike what we have come to think of as capitalism in its classic, dynamic sense. In such a society, producers would acquire accurate information about existing demands. This would lead to a highly developed division of labour, the ultimate consequence of which would be a decline in competition as each producer found exactly the niche for which they were perfectly adapted. The society would be rather technologically conservative, precisely because all the factors of production (labour and capital particularly) would be paid according to their productive contribution. As a consequence accumulation would be low, and there would be little incentive or opportunity to make massive bets on the future. Crucially, for it to function properly, a market society would require a low level of income inequality because only then would the marginal utility of money be equivalent so that prices would reflect actual demand. The great theorist of such a society is Durkheim, whose utopianism has been so often misrecognized as positivism.

Markets play some role in the allocation of existing resources under capitalism; but what really distinguishes the system cannot be described in terms of markets. If the price system were working as its most ardent proponents suggest, why would there ever be economic growth, and in particular why would it take the sharply discontinuous form that it has throughout the history of capitalism? A perfectly functional system of allocation through prices is conceivable without any growth at all; growth, indeed, would seem to require the use of resources to produce goods and services for which there is no current demand, and thus depends on breaking the link between supply and demand that the market system is supposed to preserve and regulate.

Historically markets have at best only a contingent connection to growth. There are numerous examples of market expansion leading to economic involution – the African slave trade being perhaps the most terrible and obvious. That continent paid for its participation in global markets with a massive demographic loss from which it has arguably not yet recovered. One could also mention the impact of the grain trade on East Elbian Germany from the sixteenth century which led to the consolidation of repressive labour controls on the large estates, or the penetration of the Tuscan countryside by Florentine merchants which undermined flourishing local land markets in the fifteenth century.

There have been many market societies in the past which were not capitalist and there are likely to be more in the future. The decisive point about capitalism is rather that it is a society in which the principal means of production are privately owned and in which as a consequence investment decisions are made in an uncoordinated, anarchic way on the basis of more or less blind guesses about what the future may offer by way of profits. Futures markets themselves are notoriously speculative. They do not provide actual information about the future, but at best a patina of rationality to economic activity that can only be a bet. The point, in any case, is that production oriented to existing demand plays at most a limited role under capitalism.

All of this points to a crucial problem: the striking lack of correspondence between human needs and capitalist investment. Let us again cast ourselves back to late 2022. What were the existing needs that could have been at least partially addressed by a trillion dollars of investment? Numerous candidates present themselves: the transition to green technology, expanding educational opportunities, improving access to health care . . . The fact that instead an unimaginable sum was ploughed into a technology for which, however alluring, there was no evident demand, let alone pressing need, has to count as a signal failure of capitalism, and a powerful argument for organizing the economy along fundamentally different lines.

One could respond by stressing that it is precisely this willingness to leap into an unknown future which lends capitalism its specific dynamism. But do we really need speculative blindness to discover the best path forward? There are many other sources of information – aside from opportunities to make massive profits in the future – that would provide a more rational guide for the application of the accumulated wealth of our enormously wealthy societies. The question is never asked, ‘what are the things we haven’t discovered, what are the paths we haven’t taken’ because we have had imposed on us a huge bet on something no one said they wanted and no science said was needed. The false determinism of technological progress precludes asking such questions. Perhaps the most important point to keep in mind is that history is an infinitely complex branching tree, not an ascending line – and it is politics that articulates its structure.

Read on: Aaron Benanav, ‘Beyond Capitalism —1’, NLR 153.