This dissertation examines the history of first American and Soviet teaching computers that were designed to deliver instruction to students and automate teaching. Specifically, I examine the history of the University of Illinois’ pedagogical computer PLATO (Programmed Logic for Automated Teaching Operations) and the multiple projects of the Soviet Academy of Pedagogical Sciences and the Council on Cybernetics to produce a teaching machine comparable to the PLATO computer in 1959-1980. I use these pedagogical computer projects as an opportunity to explore ways in which engineers, computer scientists, psychologists, and educators co-operated to define human learning and make it accessible to computers. The prospect of building technology that would govern the acquisition of new knowledge generated debates about the character of learning, human creativity, and whether computers could be made to understand and control the ways our minds learn. It also prompted novel discourse about the place of techno-scientific knowledge and education in the era of mass automation in the United States and the Soviet Union. Paying close attention to the software of American and Soviet teaching computers, I track how the developers of this technology built their divergent beliefs about the human mind, creativity, learning, and computation into their teaching machines. In doing so, they created new theories of learning and formal systems of human cognition that circulated across the Iron Curtain.