Last resort discovery tool for tvOS
Sometimes it’s difficult to find a movie or show to watch. After an hour of scrolling through your streaming apps without success, you may end up pulling out your phone to search for information online, only to realize you don’t even know where to start. This project is an attempt to design a “last resort” discovery tool in tvOS that allows you to reliably find something that fits your current preferences.
This tool is slow and therefore not a replacement for recommendations or regular search tools. It’s a complementary feature optimized to help you when you feel completely lost in your search for your next show to watch.
Tools: SwiftUI, HTML/CSS/JS, Figma
The final concept uses an LLM to break down content based on similarity and structures them into distinct paths. The deeper down you go, the more defined the paths are. While it’s not the fastest way to discover good shows, it’s a certain way of getting closer and closer to what you want. That’s why it works well as a last resort when other discovery methods failed.
How does it work and why does it look like it does? Check out the full walkthrough video below.
00:00
Intro
00:40
Chapter 1. This project’s relation to existing solutions
02:05
Chapter 2. Different ways of moving around
05:05
Chapter 3. Optimizing flow between rows
05:55
Chapter 4. Maintaining good overview of the options
06:49
Chapter 5. Does it work at all?
07:13
Chapter 6. Grounding in the OS
08:30
Chapter 7. Finding together
09:10
Chapter 8. Hint
10:51
Chapter 9. Ancestors
12:35
Chapter 10. Helping the system
14:48
Chapter 11. The system don’t always see things like you do
15:53
Chapter 12. Limitations of the project
Last resort discovery tool for tvOS
Sometimes it’s difficult to find a movie or show to watch. After an hour of scrolling through your streaming apps without success, you may end up pulling out your phone to search for information online, only to realize you don’t even know where to start. This project is an attempt to design a “last resort” discovery tool in tvOS that allows you to reliably find something that fits your current preferences.
This tool is slow and therefore not a replacement for recommendations or regular search tools. It’s a complementary feature optimized to help you when you feel completely lost in your search for your next show to watch.
Tools: SwiftUI, HTML/CSS/JS, Figma
The final concept uses an LLM to break down content based on similarity and structures them into distinct paths. The deeper down you go, the more defined the paths are. While it’s not the fastest way to discover good shows, it’s a certain way of getting closer and closer to what you want. That’s why it works well as a last resort when other discovery methods failed.
How does it work and why does it look like it does? Check out the full walkthrough video below.
00:0000:40
02:05
05:05
05:55
06:49
07:13
08:30
09:10
10:51
12:35
14:48
15:53
Intro
Chapter 1. This project’s relation to existing solutions
Chapter 2. Different ways of moving around
Chapter 3. Optimizing flow between rows
Chapter 4. Maintaining good overview of the options
Chapter 5. Does it work at all?
Chapter 6. Grounding in the OS
Chapter 7. Finding together
Chapter 8. Hint
Chapter 9. Ancestors
Chapter 10. Helping the system
Chapter 11. The system don’t always see things like you do
Chapter 12. Limitations of the project