The VR programming system I really wanted to make while I was at Facebook was heavily inspired by Vi Hart and M Eiffler's research. There were many other sources and a lot of first principals design but they're still some of the only people really exploring this space. https://theartofresearch.org/combination-physical-and-virtual-tools-for-spatial-computation/I like to think my system was a little more pragmatic but a lot of the "crazy" ideas that got vetoed by ux designers and engineering managers are viewed as defining or important features (non-linear execution, spatial primitives, gesture to objects). I had no AR specific ideas but I had some long term plans for extending the system to AR. It's hard to sell research to a big company product team though.