What I want to work on

July 21, 2025

---

Some of these things I don't yet know how to do. But everything I do know, I once didn't.

In no particular order:

• No, I don't want work on your payment processing system.
• Yes, I want to implement SSA form for your high performance optimizing compiler.
• Yes, I want to work on the NT kernel.
• No, I don't want to build your LLM system.
• Yes, I want to write code that will end up in space.
• No, I don't want to only write test cases for code that will end up in space.
• No, I don't want to do data processing, whatever that means.
• Yes, I want to figure out why splines made your video game 16% slower.
• No, I don't want to use React.
• Yes, I want to write GPU drivers for Linux.
• Yes, I want to dive into the assembly.
• No, I don't want to write B2B SaaS.
• Yes, I want to make new hardware.
• Yes, I want to have to refer to a specification or physics textbook before I can implement what you want in code.
• Yes, I want to use Vulkan.
• Yes, I want to create new algorithms.
• Yes, I want to change the world.

There's some flexibility. I'd still work on some of these no's, even if I don't want to. The following might make me want to:

• Lots of autonomy
• Lots of ownership
• It's open-ended (figure out the tools to do what it takes, and get the job done)
• I'm learning something new
• I won't be doing it forever