In 1951, Paul Fitts published a seminal paper allocating functions to humans or machines based on the relative proficiency for each. In the 70 years since this list was published, significant advances in the field artificial intelligence (AI) necessitate another look. In this presentation, the authors propose new guidance on allocation of functions between humans and machines in the context of AI. We begin with a background on the Fitts list and an overview of AI, and describe how AI represents a fundamental shift in human-machine teaming. We propose a new list of functions and their allocation between humans and AI, and describe the methodology used to evaluate this list of functions. And we conclude the presentation with a discussion of the implications of this list for the development and operation of AI applications, and examine how UX practitioners can leverage this information to enable effective human-machine teams.