AI, Agency, and Aloha by Anna Rezk

This year, I had the pleasure to present my work on design for user agency at the 2024 Conference on Human Factors in Computing Systems (CHI). This meant that I got to attend the largest conference in Human Computer Interaction (HCI), which happened to be held in Hawaii. Now, one month later, I still can’t quite believe I got to experience such an incredible event in such a beautiful location, and, most of all, that this is what I do for work.


Having started my PhD during the pandemic, I assumed that my chances to travel and go to big conferences overseas might be slimmer due to potential travel restrictions that seemed endless at the time. Now contrast this with being at the opening keynote of CHI, on the opposite side of the world, in a room filled with hundreds of fellow researchers in HCI.


Seeing how large the research community is – considering that the CHI attendees only make up a fraction of it – was intimidating at first (how am I meant to read all these papers?) but also reassuring (this field must be really relevant; it produces so many papers!).

An ever-growing research community


The seasonal topic was, of course, AI, with publications focusing on different implementations of it, specifically LLMs (large language models).

With over 1,000 accepted papers, it was the largest CHI conference to date, despite the competitive acceptance rate of 29%. The fairly new model of Revise and Resubmit created a two-staged process, allowing authors whose papers had potential to proceed to the next round to – as the name suggests – revise and resubmit their work based on their reviewers’ comments.


As CHI has grown over the years, the organising committee has faced increasing difficulties finding venues with large enough capacities. Combine this with their five year rota for locations – including two locations on each of the North American coasts, one in Asia, one in Europe, and one “wildcard” to reach new communities – as well as the already existing plans for the cancelled CHI 2020 in Honolulu, the 2024 cohort ended up “catching waves at CHI” in Hawaii, as the student volunteer song suggests.

Of course, the choice of location also highlighted inclusivity issues in academia, particularly pertaining to financial and visa privilege, as many accepted authors were unable to make it to CHI in person.


Conflicting schedules and the Fear of Missing Out


Once the schedule for the CHI talks had been published, I tried to identify interesting papers and plan my conference time accordingly. I was overly ambitious at this stage, causing me to double and triple book myself. This was because multiple talk sessions ran in parallel across the entire convention centre (which is so massive, it was used for shooting airport scenes in the series Lost in the early 2000s), forcing attendees to make difficult decisions of which talks to attend. I had four different approaches to that:


  • This paper might be related to my own research.
  • This paper is on something I never thought about, and I would like to learn more.
  • This paper is by one of my colleagues.
  • This paper has a funny title.


I also attended a workshop on bringing together HCI and HFE (Human Factors Engineering) at which I presented my most recent (and upcoming) work on AI-assisted agents and was able to discuss it with a group of senior academics.


Time to present


The conference days flew by until it was my turn to present on the final day, a Thursday morning. Surprisingly, and despite a tropical storm, the room filled up, and my talk was followed by multiple thought-provoking questions, inspiring new research questions I had not previously considered for my topic and now plan to revisit.

Having presented work on design for user agency in personalised news curation, the question from the audience that stuck with me the most was whether increased user agency correlated with an increase in user trust in the news. My study method – the provotype approach – was well received, and I had the pleasure of demonstrating NAIRS, my provocative algorithmic news recommender that also assesses your News Personality Type and lets you adjust it if you disagree, in the exhibition hall. To top it all off, I was awarded an honourable mention for my paper.

I was pleasantly surprised by how collaborative and inviting the HCI space is: from researchers picking up conversations during coffee breaks to others reaching out to me before and after the conference because they were intrigued by my paper and wanted to chat about it.


The takeaways


One of the most enriching experiences was exploring research fields unrelated to mine, discovering different methodological approaches, and witnessing firsthand the true interdisciplinarity of HCI. This is what best encapsulates CHI for me: a somewhat overwhelming conference, where getting lost isn’t really possible because you’ll always stumble upon some brilliant work along the way.


The steps after CHI for me were to tackle the discussion point that followed my talk as well as read a large number of new CHI papers, which mysteriously keeps growing the more I read. And who knows, maybe aim for CHI 2025, held Yokohama, Japan.