Towards a political economy of algorithmic capitalism

Abstract

The emergence of large language models such as GPT3 has prompted intense debate about the transformative potential of generative AI to ‘revolutionise’ knowledge work and disrupt economic activity. This article represents an initial attempt to problematise this phenomenon within a Marxist analytical framework. Borrowing the term ‘algorithmic capitalism’ to refer both to the broad family of generative AI models and their anticipated effects on the social form, the article offers a two-stage analysis towards a potential political economy of this new technology. First, drawing on a classical Marxist analysis of labour-capital relations, the article investigates shocks to the exchange value of labour within knowledge work which are likely to arise from the widespread uptake of generative AI. Second, the article considers the semiotic effects of generative AI, and attempts to position these within the Marxist literature on the commodity fetish via Baudrillard’s concept of the hyperreal. In so doing, the article argues that the emergence of generative AI can be also understood in terms of a further advance of the commodity fetish, in which meaning itself is reified.