哩賀(lí hó), I am Ryan Yen 
I do research, which means I spend my days banging my head against the wall trying to figure things out.
Overall, my thinking mostly revolves around AI-assisted programming interfaces. Currently, I’m exploring Semi-formal Programming at MIT VIS Lab, the a shared representation where both humans and machines can comfortably operate and co-evolve.
Works
I’ve been thinking about enable programmers to program in different ways, and in different abstraciotns.
One example is Code Shaping, where we explore using freeform sketching on and around code to edit it iteratively with GenAI, helping programmers express ideas at a much higher level of abstraction. Another is CoLadder, which explores multi-level blocks for code generation, letting users manipulate abstraction layers dynamically.
I’ve also worked outside the programming space. In Memolet, I looked at how we can reuse conversational memory in AI-human interactions. In To Search or To Gen?, we explored how generative AI and web search could complement each other in programming.
Reflections and Repentance
I spend many time reflecting on my thoughts… and I like writing them down. If you’re curious about the things I’m noodling on, you’ll find more on my Blog.
Readings
A lot of my thinking is shaped by what I read. Honestly, reading feels like time-traveling conversations with other thinkers. Nerdy, I know, but it’s true. I just wish I had more time to do it.
Some of the pieces that have stuck with me, including Reification, Polymorphism and Reuse, Demonstrational Interfaces, Cognitive Engineering, Designing Interaction, not Interfaces, Instrumental Interaction, Direct Manipulation Interfaces, Thinking with External Representations, Ultra-Lightweight Constraints, What Is Interaction?, and Usability Analysis of Visual Programming Environments.
Some News
- I am currently intenrting at Microsoft Research, working with Steven Drucker and Chenglong Wang
- I got a paper at DIS 2025 on To Search or To Gen?, working with Nicole and Jian
- I received CHI 2025 Best Paper Award for Code Shaping