What are you reading?

What Are You Reading? Daniel M. Russell

Everything Everywhere All at Once isn't just a movie title; it's also a spot-on description of what's happening in the human-centered AI (HAI) literature.

HAI as a field is evolving rapidly. Systems are being built and launched at a record pace with Internet-circulated, breathless posts summarizing the very latest developments and innovations ("The 1,000 Prompts You Must Know"). What is clear is that established publication venues are not able to deliver the kind of content and insights that developers, designers, and researchers need to react in real time.

To be clear, books are wonderful cultural artifacts, but the time between initial idea and publication is usually measured in years. Print journals run to months—often many months. Even popular conferences operate on a pace of several months to almost a year. What I am saying is that the traditional ways of staying in the know on a topic that is evolving this quickly do not work well.

So, what's an Interactions reader to do? I polled 20 of my colleagues and practitioners who work in the HAI area about what they're reading and how they're staying abreast of this rapid pace of development. It's not a surprise to learn that everyone's focus has turned from traditional publication resources (books, journals, conferences) to the less vetted (but very fast turnaround) sources such as papers on arXiv or posts on Medium, Substack, Mastodon, and LinkedIn. People are turning to places where updates occur daily—even a weekly news publication tempo now feels somehow quaint and relaxed.

Here's what I'm doing. First off: I'm paying close attention. Things are changing rapidly. This is the time when new paradigms are emerging, shifting, and transforming and at a pace we did not expect. It is clear that everything currently being published is provisional. I'm remaining skeptical. I'm reading high-reputation news outlets, emails, and email digests from experts. I am looking to friends and colleagues for recommendations about what to read and what trends need attention—and what trends do not. I'm paying that social capital back with my own recommendations. I recommend investing time to read what is out there across all the platforms to which you have access. Yes, it's growing and shifting rapidly, but it is worth knowing the players and who is setting the tone, crafting the story, and hoping to set the rules and governance around AI and HAI, including companies, governments, teams, and individuals.

I admit that this is a new mode for me; I mostly ignored these channels because, historically, I could predict what they'd say. I can't predict accurately any longer. For the first time in a long time, these "popular" sources are the bleeding edge of reflection and insight when it comes to the technology industry and HAI.

For the first time in a long time, these "popular" sources are the bleeding edge of reflection and insight when it comes to the technology industry and HAI.

Is this the end of the role for books, journals, and conference papers? Probably not. We shall see what books, and what conference and journal papers will make an impact in the next few months to years. In the meantime, here is what I am reading right now. This list is neither exhaustive nor complete; these are what have proven useful for me.

ArXiv: There are several arXiv resources. One to attend to is https://arxiv.org/list/cs.HC/new

Medium: https://medium.com/search; search for keywords of interest

Reddit: https://www.reddit.com; look around, lots of resources

Universities with active HAI publications:

• https://hai.stanford.edu
• https://hcil.umd.edu/human-centered-ai
• https://www.hcii.cmu.edu/research-areas/human-centered-ai
• https://www.uu.nl/en/research/human-centered-artificial-intelligence

A fairly complete list of HCAI programs worldwide: https://hcai.site/groups

Newsletters that I currently look at include:

• MLearning.ai (https://medium.com/mlearning-ai)
• Ben's Bites (https://news.bensbites.co)
• Human-Centered-AI (https://groups.google.com/g/human-centered-ai)
• The Batch (https://www.deeplearning.ai/the-batch)
• Inside AI (https://inside.com/ai)
• The Road to AI We Can Trust (https://garymarcus.substack.com)
• On Tech: A.I. Newsletter (https://www.nytimes.com/column/on-tech-ai-newsletter)
• The European AI Newsletter (https://www.europe-anartificialintelligence.com)
• ChinAI Newsletter (https://chinai.substack.com)

  Author

Daniel M. Russell was a longtime user experience researcher on Google's Search team. He is now a free-range researcher splitting his time between the computer science departments at Stanford University and the University of Zurich, where he teaches courses on the intermingling of HCI and AI. [email protected]

Copyright held by author

The Digital Library is published by the Association for Computing Machinery. Copyright © 2023 ACM, Inc.