Approaching a new field
How do you approach a new field optimally? I am exploring ml systems in my free time these days and had to answer this question for myself. The typical method would be to find the people working in the field and question them. Go through opensource codebases and crawl through some Pull Requests. Lurk around the discord channels and gauge the vibe.
Read some of the latest papers in the field, they won't be understandable yet but the point like everything else is to do a vibe check. How open are the people at the cutting edge with sharing their research and work? What is hidden and what is out there? What is the industry-academia split?
The most important thing though is to have the right mindset. This is something I have realised. It's important to have an open mind and cultivate curiosity. To explore. I always find that entering a new field with fixed goals limits your perspective and often invites disaster.
In the beginning, you know very little to really understand what you want, what is out there, or which parts pass the vibe check for you and which don't. It's easy to get discouraged by the mountain of knowledge you need to climb or to get tunnel vision early on and miss out on what really matters to you.
This is where you need to have faith in yourself. Trust yourself enough to explore. Believe that you will find your way and establish a process. After that, it's just about trusting the process. Like anything it is just work and some faith. Faith in yourself, a higher power, or experience of kind strangers on the internet :)