Field Notes 2/14/26
This year, I decided to spend my Valentine's Day trying something new -- a hackathon. I was inspired by a Youtube video about analogue switches people were making to lower cortisol and overwhelm (the world is a lot right now). I wanted to create music player that captured some of the charm of analogue (ironic? yes). Oftentimes, when people think of analogue music, their mind goes to record players. However, as a 90s child, what feels most nostalgic to me is the CD player.
With just a few hours to design and build, I wanted to keep the project bite-sized. In the spirit of analogue, I wanted the controls to be limited, the UI to be skeuomorphic, and the focus to be on listening to an album in order, all the way through.
For my tech stack, I used Codex, Figma Make, Spotify Developer API, Github, and Vercel. By the end, I achieved a working prototype (yay!) which I published on Vercel. However, there are many limitations -- due to limitations of the free tier of the Spotify Developer API, only people I manually add can use this site (max 5). Also, the interface lacks a lot of finesse. The interface is limited (as intended), but continuing to enhance the skeuomorphism through UI elements (texture, 3D buttons, etc.) and sound would enhance the feeling of play in this interface.
Overall, I approached this project as play. I did not set out to take home gold. Instead, I wanted to try to work differently than I have before, stretch my abilities, and get outside my comfort zone.
In the process I learned a lot. First, I found that when collaborating with an AI tool, starting with a clear intention keeps you focused, but since you are working with a non-deterministic system, you need to remain open and riff with the tools. Where I see design coming into this process is recognizing what's worth keeping (even if it was unexpected) and what should be thrown out (even if that was part of your original brief). Speaking of tools, the other thing I discovered was the value of shopping around. Right now, AI tools that can code or build or generate visuals or text are plentiful and the barriers to switching are very low. When you are starting out or when you hit a wall, switching systems can be a way to get unblocked. Another way to get unblocked -- try switching inputs. When written prompts aren't cutting it, a sketch, a mood board, or a reference image can push things in the right direction.
Overall, my takeaway was that it is still very early. As we all know, the tools and models are changing fast. The implications of these and other AI tools for our society, and world, are sweeping. In the face of all this upheaval, it can be tempting to shut out the noise or bury your head in the sand. But, being able to shape the future we want to see requires getting your hands dirty, playing around, and gaining our own understanding what's out there and what we can do with it.
This year, I decided to spend my Valentine's Day trying something new -- a hackathon. I was inspired by a Youtube video about analogue switches people were making to lower cortisol and overwhelm (the world is a lot right now). I wanted to create music player that captured some of the charm of analogue (ironic? yes). Oftentimes, when people think of analogue music, their mind goes to record players. However, as a 90s child, what feels most nostalgic to me is the CD player.
With just a few hours to design and build, I wanted to keep the project bite-sized. In the spirit of analogue, I wanted the controls to be limited, the UI to be skeuomorphic, and the focus to be on listening to an album in order, all the way through.
For my tech stack, I used Codex, Figma Make, Spotify Developer API, Github, and Vercel. By the end, I achieved a working prototype (yay!) which I published on Vercel. However, there are many limitations -- due to limitations of the free tier of the Spotify Developer API, only people I manually add can use this site (max 5). Also, the interface lacks a lot of finesse. The interface is limited (as intended), but continuing to enhance the skeuomorphism through UI elements (texture, 3D buttons, etc.) and sound would enhance the feeling of play in this interface.
Overall, I approached this project as play. I did not set out to take home gold. Instead, I wanted to try to work differently than I have before, stretch my abilities, and get outside my comfort zone.
In the process I learned a lot. First, I found that when collaborating with an AI tool, starting with a clear intention keeps you focused, but since you are working with a non-deterministic system, you need to remain open and riff with the tools. Where I see design coming into this process is recognizing what's worth keeping (even if it was unexpected) and what should be thrown out (even if that was part of your original brief). Speaking of tools, the other thing I discovered was the value of shopping around. Right now, AI tools that can code or build or generate visuals or text are plentiful and the barriers to switching are very low. When you are starting out or when you hit a wall, switching systems can be a way to get unblocked. Another way to get unblocked -- try switching inputs. When written prompts aren't cutting it, a sketch, a mood board, or a reference image can push things in the right direction.
Overall, my takeaway was that it is still very early. As we all know, the tools and models are changing fast. The implications of these and other AI tools for our society, and world, are sweeping. In the face of all this upheaval, it can be tempting to shut out the noise or bury your head in the sand. But, being able to shape the future we want to see requires getting your hands dirty, playing around, and gaining our own understanding what's out there and what we can do with it.
