Data are not neutral but reshape reality

Contemporary economic empires are founded on data. Google and Meta are classic examples. But also, in different ways, Tesla, Caterpillar, or General Electric. Digitalisation multiplies the availability of data, which makes the advent of artificial intelligence possible, which in turn characterises the new evolutionary phase of digitalisation. All of this is happening thanks to the growth of the computing capacities of increasingly voracious data centres, in the context of an unprecedented concentration of economic power: founded, precisely, on the control of enormous quantities of data.

But what, in the end, are these data? Why are they such a strategic resource, even though they are not in themselves scarce? How do they multiply the value of the companies that possess them? And what are the rules of the system that produces, preserves and enhances them? Cristina Alaimo of the École Supérieure des Sciences Economiques et Commerciales and Jannis Kallinicos of Luiss address these questions in their book 'Data Rules. Reinventing the Market Economy' (Mit Press 2024). And they go so far as to theorise a new evolution of essential market structures.

"When facts become data, in digital format, they become separated from the context in which they happened," Kallinicos tells Nòva. "They become elements of a knowledge in the making that transforms the economy, society, culture." The knowledge economy needs digital infrastructure and this is essentially a system for managing data. Which are born from facts but, when recorded digitally, take on a different substance. In Kallinicos and Alaimo's reconstruction, data are described as 'ephemeral', 'malleable', 'editable', 'executable', 'performative'. Facts exist in their historical and social context and are exhausted there, while data live in an abstract, operational, not merely technological, but 'socio-technical' dimension. In the platforms that collect them and the algorithms that process them, one sees relationships between data emerging from different contexts that can generate correlations, regularities and knowledge that would otherwise be impossible to recognise. Which among other things also enables the creation of powerful generative artificial intelligences.

"Data are agnostic to the context that generates them and are vital to the development of universal machines such as computers, attentive to the efficient processing of data but not its meaning" as Kallinicos puts it. "But they are not neutral: in the sense that the choices that produce and value them are relative to the value systems, the cultural and institutional histories, of those who designed the ways to collect and use them."

Data analysis, therefore, is not merely a matter of statistics and technology. "They do not only derive from the convergence in digital of different economic sectors. But they certainly generate a cultural homogenisation: because any human experience can be dataified and become comparable to any other".