This study asserts that nineteenth-century mnemonics, known as ‘artificial memory systems’ until the twentieth century, served as one of the cornerstones for the invention of modern computing. Mnemonics systems rejected the associationist conceptualization of an intertwined mental hierarchy of ideas. As I show, Victorian mnemonics systems instead conceptualized a new form of memory akin to random access memory; they developed systems in print that perform as algorithms; and they invented universal and executable languages before programming. The roots of this project lie in my discovery that, unlike their predecessors, Victorian mnemonic systems were realized as autonomous memory machines. “Autonomy” here is defined not by physical externality from humans, but rather by a random-access recall process that consists in a printed system worked partly by the mind. The story that flows from this insight that Victorian mnemonics devised an autonomous memory system revises the standard accounts of technology historians and digital scholars and challenges the consensus that mnemonics usage declined in the eighteenth century to reposition Victorian mnemonics as an exciting new dimension of the history of computing.