what am i doing?

i’m nearing the end of my phd. was i happy doing it? for large parts of it, the answer has been no.

some of that has just been old patterns that have pushed me towards doing work that superficially feels nice, but is really avoiding my emotions and leads to stress, anxiety, frustration.

but other parts of that have been trying to change my work process to be someone i’m not. i need to be more example driven. i need to work on real problems. i need to build real systems. there’s something about the way i’ve received that kind of advice from other people, from the world, from myself, that has pushed me away from what i really enjoy doing. i can feel like i’m fighting myself and i’m pushing against something. i’m swimming upstream.

i’ve been working my way through an algebraic topology lecture the past couple days and it’s just so fun! not in a way that feels like i’m avoiding things. in a way where it just feels nice to study a beautiful theory that formalizes an intuition. and one where simple diagrams help my intuition and i can see that happening in real time

this kind of math is a kind of art. the kind of art that instills awe in me, that brings me closer to the divine. applications are not the primary focus

i didn’t realize how disconnected i’d gotten from studying and finding these kinds of beautiful theories and finding ways to connect them to intuitive understanding. a lot of what i was obsessed with in undergrad was finding those things. it’s what drew me to dependent type theory and infinitesimals and synthetic differential geometry (not that i understood any of those super well). i could see how mathematical foundations were a choice. and how they were formalizing intuitions and logics and measurement. that was really cool. that is really cool

i’ve been unhappy coding gofish recently and i was trying to understand why. i thought that moving from the lower-level bluefish system to something higher-level in gofish, where shorter programs are more interesting, would help me enjoy things more. but i feel almost as unhappy as last year. (and actually i have moved away from diagrams, which were always so fun to me!) i started thinking about which parts of the process i enjoy and which i don’t.

i like formalizing “underlying space” and wrapping up the low-level api into the mid-level api. i like finding more detailed representations of data types. (although even those parts have felt fraught lately). i don’t really love picking the best names for things. i don’t love doing the python wrapper.

i think i have found ways to integrate examples and user research into my process. and i’ve found ways that they ground me. but maybe at the end of the day i’m not super duper interested in spending most of my time on the “surface” level design. i love the structures. and i love the beauty of those structures for their own sake.

visualization is a really interesting domain for studying structure, because it lies right at the boundary between formalisms of space and formalisms of perception. metrics and open sets capture notions of proximity, connectivity, etc, but they don’t quite match how we actually perceive things in graphic representations. they don’t capture how we use scatter plots and pie charts to make comparisons. and so we might need new structures to understand those phenomena

and maybe this is what’s making it hard for me to find out where to do my work. because right now the vast majority of research in HCI and VIS is less concerned with structure and more concerned with surface-level design. it’s very hard to publish work on structure, and reviews don’t tend to push authors to think more about structure. other subfields seem to value structure more. i want to be free to do “basic research.” i want to work in a community that makes the weird, exploratory prototypes or the deep systems work on HCI, VIS, and programming interface problems. and I want to be grounded in users’ perceptions and expectations… at least a little bit

i like structures