Is This Thing Still On?

Struggling is Striving.

I have struggled recently in where and what I want to do with my life. I always thought I had this weird disease, where if I felt like my life was planned out, then I get really depressed. I only ever felt right when I was in a position I didn’t belong. Where I needed to prove my worth with my feet held to the fire. When the path forward isn’t set, when you aren’t guaranteed victory, that is where I find my true self.

I finally pieced together that it isn’t that I dislike comfort or an easy life, I dislike having no progress. I cannot accept the easy path for it means I am done progressing, that my upward mobility is done. I have lived a life of relative leisure, and with that you are given a chance to stay in your lane for a “better than most” lifestyle. In this, I have yet to find an external ceiling, a barrier to entry I cannot overcome. My biggest enemy is my own contempt.

It is in this that I have decided to take larger leaps on greater problems to really try and find my external ceiling. My delusion tells me there is no ceiling, for all is solvable given enough time. I have chosen machine learning for it seems to be a feat that is challenging with a lot of prerequisites(most that I don’t have), but also because it is a path walked before. A guiding rope from those before, to help me catch up to the pack. Then hopefully in this new frontier I can find my ceiling, suffer my inevitable ego-death, and no longer be depressed for I have reached my limit.

Goals

All of the above just to say, I want to learn as much as I can, as deeply as I can about GPU Engineering. My current measurable goals that are subject to change as I actually get into it are:

• Develop a strong foundation in machine learning and reinforcement learning.

• Learn modern model architectures by implementing them from scratch.

• Build an intuitive understanding of training dynamics and representation learning.

• Read and implement current machine learning research.

• Publish progress and technical writeups at least once per week.

• Contribute to open-source projects where it meaningfully overlaps with this work.

• Long term: make ML engineering my day job and contribute work I’m proud of.

“Difficulties strengthen the mind, as labor does the body.” —Seneca