Instruction

Instruction explores the empty terrain of aspiration. The internet contains countless videos promising salves for desires both small and large — “how to successfully complete projects,” “how to have fun at home,” “how to love your busy life.” Each video offers confident platitudes, urging viewers to consume some product, embark on some self-care fad, and manifest contentment through persistence and faith. Instruction uses custom software to piece these videos together, iterating through shifting faces from sources both real and synthesized, forming a chorus of cheerful intonations blurred together into one. The disconnect between the relentlessly positive speakers, flat captions, and unnerving imagery is meant to be disorienting and uncomfortable, echoing the complexity and abjection of capitalism, climate change, and social dissolution. The fact that we want answers to these questions reveals something fundamental about how modernity has failed us. The answers on offer are not sufficient.

Instruction uses open-source neural networks to achieve constantly-changing, nonexistent identities for the speakers pictured. The use of AI here is deliberate, transforming each frame into an imagined aggregation of humanity. Thus the empty stares of Instruction’s subjects reflect a digitally modeled personhood extracted by algorithms through immense quantities of data — yet this representation remains uncanny and wrong. These faces aspire to be everyone, yet they are no one. The neural networks underscore Instruction’s central contradiction, how a barrage of manufactured content aimed to assuage real, human aches only further alienates.