about

Xerox. Children using the Alto personal computer, ca. 1979. Courtesy of PARC.

Xerox. Children using the Alto personal computer, ca. 1979. Courtesy of PARC.

Motivation

I've long been fascinated by the rich history and academic research behind human-computer interaction and interaction design. As a designer, I'm always seeking to learn new things that can fuel my work with fresh inspiration and perspectives. I often find that reading papers provides me with meaningful depth and knowledge that appeals to my innate desire to learn more about the world, and this field that I am so passionate about.

Over the years, I've realized that academic literature isn't always the most accessible, especially for my fellow designers who've arrived in this field through their own unique, non-traditional journeys. Papers can be dense, technical, and frankly dull to pore through sometimes. With so much research out there, just figuring out where to start can be an intimidating barrier.

My hope with this website is to synthesize some of the seminal HCI papers into simplified outlines and topics. This is by no means a comprehensive academic guide, but rather my humble attempt to distill the key developments in HCI history into digestible takeaways. A little passion project from one learning designer to others who share my curiosity.

Through this effort, I learned so much about the foundations of HCI and also became acutely aware of all that I don't yet fully grasp. But most rewarding of all, was discovering so many fascinating perspectives that I believe will help me approach design challenges in new and interesting ways.

If you found any value in these amateur compilations, or they provide a springboard to explore papers that pique your interests - I'm overjoyed! My sincere wish is that this site sparks your own learning adventures into the captivating world of HCI. A little vantage point for you to explore your own interests in this wonderful field.

Looking Back, Looking Forward

With AI ushering in a paradigm shift, it's especially important today to bridge the academic, and practitioner sides of HCI. As Susanne Bødker notes in HCI Remixed, "..we need to keep reminding ourselves of how and why, our everyday technology came into being. In order to be better at pointing to the future, it needs to be aware of its history too."

I really believe that understanding our academic history is crucial for innovating thoughtfully. While we must keep sight of future opportunities, we shouldn't overlook the hard-won UX principles of the past. In fact, longstanding HCI concepts, once impossible, may open new possibilities when reimagined through modern capabilities.

If my little contribution helps even one fellow traveller uncover inspiring ideas - that alone makes every hour invested here worthwhile.

Process & Tools

I built this site using Next.js for the frontend framework along with Tailwind for styling. For the backend database, I went with Supabase for its simplicity and generous free tier.

Figuring out where to start with all the seminal literature was definitely intimidating. I leaned heavily on AI to help point me in the right direction. I began by prompting OpenAI's latest natural language model, GPT-4, to suggest some of the most influential HCI papers over time. I supplemented the AI's recommendations using tools like Connected Papers to discover additional relevant works cited together. Reading lists from top university programs , alongside references from books such as Ben Shneiderman's Human-Centered AI also provided a great foundation.

To generate the paper summaries, I created GPT-4 Assistants on OpenAI and simply prompted titles without any full-text uploads. The impressively accurate overviews demonstrate how much knowledge these models have accumulated through pre-training alone!

Organizing the sprawling academic content coherently was an iterative learning process. After several attempts plus many GPT-4 prompting cycles, I finally landed on categorical groupings that made intuitive sense. The interconnected and overlapping nature of HCI research makes it quite challenging to categorize papers into neat buckets, but I eventually landed on the current structure, which feels right, and provides a nice overview.

I also implemented semantic search using a Retrieval Augmented Generation (RAG) with a Pinecone vector database backend. I created embeddings of the the paper summaries which are then used for comparison against user queries, surfacing relevant results. Ultimately, interpreting search queries by semantic similarity rather than just keywords. The results are decent but could be further improved by higher quality embeddings from larger volumes of HCI content, or enhanching the search through keywords. There is ongoing research in this area, and as capabilities and techniques evolve, I hope to integrate them into this project.

On the whole, this project gave me hands-on experience with cutting-edge AI across summarization, classification, information retrieval, and language generation. I also gained full stack web development skills thanks to Next.js and Supabase. But most of all, it's thrilling to have a platform to share my passion for HCI, and to be able to contribute to the community in my own small way.