Vector Theory
David M. Berry "There is no software" Friedrich Kittler, 1995. Figure 1: A vector and its mapping into "vector space". Kittler was right, but, perhaps, not in the way he intended. When he declared that software dissolved into hardware operations of voltage differences, he was making a materialist claim about symbolic computation, about computer code that could, in principle, be read, parsed, and followed through logical gates. His analysis traced how each layer of software abstraction conceals the operations of the layer beneath it, an opacity that serves power, but the material substrate he traced still operated through discrete, Boolean logic, through states which could, in theory, be read. In the 90s, the "hidden" layer was the voltage. Today, the hidden layer is the sub-symbolic weight structure . What Kittler could not have anticipated was that computation would escape the symbolic paradigm altogether. However, the weight matrices of contemporary AI r...
