<aside> 🧩 Lisa Leung is a junior in Information Systems and Human-Computer interaction. Her passions stretch wide and far as she’s interested in consulting, design, and technology as well as systemic design. In the past, she’s been a Design and Development Volunteer at re:Bloom, a Technical Writing intern at Amazon Web Services, and an Undergraduate Research Assistant at CMU!
</aside>
The introduction
What got me really interested in HCI was actually my primary department, information systems, and through one of the classes that we were able to take: Mobile Web Design and Development (67-240), with Professor Moussawi. We spent the first half of the semester mostly learning about HTML, CSS, and JavaScript, but then also wire-framing and usability, and that's where we got really into HCI.
The exploration
With freedom of electives and through some of the core classes, you can take on different roles and team up with students of diverse background while still being in one cohesive program.
You can learn about usability heuristics, be hands-on in development, and solicit feedback through user research. There is lots of freedom in the course and support from the instructors.
AI as a Design Material
This is neither a tech-heavy class nor a design-heavy class. But instead, it is ideation-heavy. It's not about building out the coolest AI you can or designing the most beautiful user interface. Instead, it's about understanding AI's affordance as design material, recognizing its foundational functionality & limitations, and coming up with creative ideas to exploit low-level technology that brings high-value solutions.
An example: Google Translation
One ideation technique is to match groups of users to the technology rather than the other way around. In other words, matching pre-existing technology and pre-existing users instead of innovating new technology.
Have you ever tried translating something with Google Translate? Let's say we're translating from English to Chinese. Grammatically speaking, Google Translate is bad. However, it does a great job when translating word by word. So that exposes both a strength and a weakness of NLP. The instructors guide students to realize we might not necessarily have a direct translation from Spanish to Farsi and have people understand it. But we can take the fact that AI can tokenize words. So let's see where we'd find that technology being used.
The skill to match-make
The instructors not only teach how to apply AI-related “design material” like Computer-Vision, Natural Language Processing, and Machine Learning. But they also introduce generalizable theories that you need to incorporate any emerging technologies and make them into something useful.
How might one practice ideation on their own
A common method in HCI is Crazy 8, where you have 8 minutes to come up with eight different ideas, and after those eight minutes, you get to go through with your team to discuss them.