12 years ago, wrote this classic rant on the challenges of doing systems work in academic human-computer interaction: dubfuture.blogspot.com/2009/11/i-give
I know this post generated a lot of discussion and intentions. I’m curious—how’s the situation different now vs. 2009?
Conversation
i'd also be curious what the state of things are now. it seems like clean, well-defined, "bite-sized" problems is a phenomenon in all of academia though, rather than specific to HCI...
2
1
3
i'm tempted to say it's a feature, not a bug (better to do real-world systems work in industry, interfacing with real users for a sustained period of time). are there reasons to shift more of this to academia?
2
I think put it well in the comments section: "systems with no present market but with the potential to create or change one.” CSCW is a generative recent example.
1
1
3
interesting, i love this breakdown. one thought here: if case (3) applies, what does full-fledged system-building get you over building an "atomic" proof-of-concept? e.g. if the market is that nascent, maybe we need to explore smaller test cases first before something e2e?
2
e.g. maybe system-building is useful only when most of the fundamental principles are already there (not sure how much i believe this)
1
I think it’s awfully entangled. Your system reifies some powerful initial insight, but building and observing the system helps you improve your insights, which generate improved systems, etc. See “How to invent Hindu-Arabic numerals?” in that piece for more on this.
Show more replies