The AI as an envirotechnical system

Increasingly, computational systems such as various applications of artificial intelligence (AI) are recognised as sociotechnical systems, sociotechnical ensembles, figurations, and alike. This presentation will argue that it is necessary to go further. Despite the benefits of Actor-Network type of models, social constructionist approaches, as well as most economical models, remain oblivious to certain kinds of materiality and related effects. Following the concepts of _envirotechnical system_ and _envirotechnical regime_ proposed by Sara B. Pritchard, I argue that all technologies start with the environment and are fundamentally grounded in various forms of human-environment relations and mutual effects in that sphere. Not restricted to resource extraction, observing AI as envirotechnical reveals how AI systems intertwine and interfere with cultures, societies, human habits and meaning-making practices. The extracted value from global commons — material and semiotic alike — is objectivated as computationalised data. Meanwhile, the discursive portrayal of AI as a kind of pre-conscious creature overlooks its unavoidable networked materiality on a planetary scale. The envirotechnical regimes are looking for ways to legitimate, naturalize, and cater to their projected future imaginaries of AI. Perhaps inadvertently, they mimic the use of the _miraculous_ or automated temple technologies as means to legitimate the ruling power dynamics in Ancient Greek society. In so doing, AI systems extend into the peripheries of thought, reproducing habitual colonial Eurocentric patterns.