My PhD thesis about my ongoing attempts to marry AI and HCI is now up. (A lot of it is about academia, where those two fields do not like to talk.) Plenty of stuff about paper tools, rapid visualization prototyping, and the connection between domain-specific languages and direct-manipulation interfaces. https://escholarship.org/uc/item/98c6j416#article_main

Looks like great work, thanks for sharing! There should be more attempts like this at framing problems tackled in disparate scientific communities with unified conceptual frameworks. Have you heard of the theory of instrumental interaction by Michel Beaudouin-Lafon? It aims at generalizing the concept of direct manipulation interface, through that of "interaction instrument" that mediates between users and domain objects. I think this is quite relevant here, although the concept of interaction instrument may have a narrower concern than that of software instrument.

It is new to me, but I see I have retreaded some of HCI's ground, Thank you! I'd come across the Document Lens, which is one of their case studies, but not within the appropriate generalization to non-textual domain objects. The scope seems very similar; for example, navigation and filtering widgets are two familiar categories. The major difference seems to me that the authors wrote in a world prior to D3.js and its applications to (say) plotting grammars. That in turn brings us to the end-user programming problem - whether users, even expert users, want to learn a new language, - which is what motivates me to use CSCW's lenses of the paper tool and the boundary object.

glad I could bring a useful reference 😁 and I'll definitely give a more careful read to your thesis! (just read the intro for now)