This “reflections” piece identifies major themes emerging from the commentaries in the issue. I discuss how some existing issues with the levels of automation (LOAs) approach may be further resolved as well as how different perspectives can lead to richer human-automation interaction (HAI) model development. The overarching objective here is to offer some reflections and last thoughts on the overall discussion of LOA issues, but of course, there may be some concerns on which some researchers may need to “agree to disagree.” Related to this, there are a few observations in the commentaries that I respond to with points of fact. Lastly, I suggest a set of future research directions based on several methodological recommendations by commentators as well as my own perspective on what is needed in HAI modeling to move forward.