At the Johns Hopkins Applied Physics Lab (APL), we use a variety of approaches to allow and encourage staff to embrace the efficiency gains possible with burgeoning AI capabilities. Different groups within APL have used various approaches to train Research Software Engineers (RSEs) on AI-enabled workflows. We describe how lab-wide policies, access to cutting edge models, and training options have enabled staff to experiment with and learn, even as the cutting edge advances rapidly. We provide some brief use cases to illustrate outcomes.