This study explores the use of deep model predictive control to overcome limitations in throughput and dynamic control for single-cell gene expression. The researchers demonstrate the accuracy and precision of using deep learning models to control gene expression in thousands of cells, including the ability to tightly control expression levels in single cells. They also apply this control to the expression of an antibiotic resistance gene, providing valuable data on the relationship between expression levels, growth rate, and survival. The study highlights the potential of deep model predictive control for driving gene expression dynamics in various experimental contexts.