Toward an Energy Theory of Value

ABSTRACT. In the ‘knowledge economies’ of the twenty-first century, the apparent supersession of the factory by the so-called ‘immaterial’ algorithm has resulted in a crisis of conceptual modeling such that alternatives to Marx’s labor theory of value have been proposed which are not easily distinguishable from subjective notions of price or affect. These affective theories correctly observe that human labor power, considered through Marx’s frameworks of absolute and relative surplus, can no longer fully represent contemporary value production; we must accept that non-human forms can also produce value. If the labor theory needs a replacement, however, we should be attentive to the utility of its groundedness in materiality, which provided the basis for a critique from outside the subjectivities produced by capital. I argue that any theory hoping to assume the analytic function the labor theory once provided should have continuity with its commitment to material conditions. This talk considers the extent to which digital and microbiological assemblages’ production of effects that are abstracted or differentiated from human labor trouble, on the one hand, the applicability of the labor theory, and on the other, the validity of the ‘immaterial’ or ‘affective’ theories that have sought to replace it. Special consideration is given to a comparison of ‘generative’ machine-learning operations to effects produced by microbial forms, each of which, I contend, constitutes value production beyond what is describable as ‘general intellect’ or ‘dead labor’. At the same time, the techno-utopic rhetoric surrounding such algorithms largely ignores their continuous reliance on material extraction. Fundamentally, the talk seeks to challenge the characterization of algorithms as ‘immaterial’ and recast them as ‘energetic’ in order to bring them into conceptual commensurability with value production by human and non-human lifeforms, and to situate computation within a framework of environmental cost.