We study how advances in automation technology affect the division of aggregate income between capital and labor in the context of long-term growth. Our analysis focuses on the fundamental trade-off between the labor displacement effect of automation and its positive effect on productivity in a basic task-based framework with an automation price schedule for state-of-the-art tasks. We get terms and conditions for automation technology and engineering change driving automation to displace labor. We identify a unique task technology that reconciles the Kaldor facts with the presence of automation along a balanced growth trajectory. We show that this technology aggregates to the Cobb-Douglas production function, thereby providing new task-based microfoundations for this functional form of workhorse. We use our theory to study the link between recent declines in the labor share and the unique nature of the current computer-fueled wave of automation.