The rush to run a user study or need-finding interview in today’s HCI derives from an under-appreciation of theory.
Remote
Ian Arawjo
@ianarawjo@hci.social
mastodon 4.7.3Incoming Asst Prof at Université de Montréal (Jan 2024). Postdoc @harvardhci@hci.social; PhD @CornellInfoSci. Programming and culture, notational programming, AI and HCI, critical computing. Building ChainForge (chainforge.ai). Prev: intern, AI/ML research.
180 Followers
102 Following
14 Posts
Joined November 12, 2022
ChainForge:
Personal Site:
Reviewing papers. Folks: if you write a Related Work section and your work applied "AI"/LLMs to some well-known domain, you are not excused from engaging with all other "non-AI"/"pre-AI" work in HCI! 😑
Voilà --it's a start! https://hci.iro.umontreal.ca
Want to research how multi-modal AI will transform programming? I’m recruiting! Prospective PhD students, email me, or apply via DIRO or MILA-DIRO affiliation: https://diro.umontreal.ca/en/english/programs/graduate-programs/phd-in-computer-science/admission/ Please re-toot and share! 🥳
Accepted (conditionally)!! #CHI2024
Wrote up some thoughts about a growing pattern of "contrived controls" in HCI systems research. What am I talking about, and why should we be concerned? Read more here: https://ianarawjo.medium.com/kicking-the-leg-out-from-the-table-on-contrived-controls-in-hci-systems-research-9f77b28fef39
Hello elephant network! I am on the job market this cycle!
My work sits at the intersection of programming and HCI. I study programming as a social and cultural practice, and design new tools+activities based on this understanding. Future directions include AI and pen interfaces. For more details+contact info, check out my site at: http://ianarawjo.com/
RTs appreciated! 😇
Booked for PLATEAU 2024! See some of you at UC Berkeley on Feb 19-20 to chat programming + HCI! 🥳 https://2024.plateau-workshop.org
Replying to
@zbucinca Yes, and I've found they tend not to engage much, if at all, with HCI literature. Arrogance, ignorance, or both?
A pre-print of our paper, “ChainForge: A Visual Toolkit for Prompt Engineering and LLM Hypothesis Testing,” is now up on arXiv. This is research with Elena Glassman, Martin Wattenberg, Chelse Swoopes and Priyan Vaithilingam: https://arxiv.org/abs/2309.09128
Hey folks! We're looking to interview developers who've used LLM API(s). We want to understand users' experiences with AI model updates (e.g. OpenAI updates). Virtual interviews will take 45min and receive $30. Sign up here: https://forms.gle/CTVCjgoJvwRhqUxr5
Has anyone tried ChainForge and then stopped using it, because it didn't seem to address the problem you had or was too hard to use? Curious about whether and why! --lmk your experiences here or in a DM!