We launched artificial intelligence (AI) control of turbulence as early as in 2017. The early control system we developed includes four essential components, i.e., a control plant, a sensing unit, an execution unit and a controller whose core component is a machine learning algorithm. The control enjoys great success when being deployed for enhancing turbulent jet mixing and manipulating turbulent flow separation from a curved ramp and boundary layers. The experimental work was further expanded to AI-based virtual engineering technology for aerodynamic shape optimization and active drag reduction of terrestrial vehicles, underwater vehicles and very large crude carriers. In all these cases, AI control, be passive or active, has vastly outperformed conventional techniques, and the huge amount of data generated out of its learning or optimization process further allows the findings of scaling laws and new control mechanisms.