First post!!⌗

eyo wuz good

I’ve decided to start a blog to document (in depth) most things technical I learn over the next ~2 years. This decision is driven by two primary motivations:

(a) it holds me accountable by making everything public. As Andrej Karpathy has aptly said, “fear of shame increases the quality of your work.”

(b) it forces me to deeply understand the things I learn in order to explain it back clearly to an audience (the existence of which cannot currently be verified)

The scope of topics on this blog for now will be systematic trading and semiconductor chip manufacturing. Very STC because I might pick up something new spontaneously or decide to write about a broader range of topics. In any case, you can expect at least these two themes to be reoccuring for the next several months. So, couple of elephants in the room here:

Why semi? I recently accepted an offer to research ML applications in semiconductor design/reliability predictions at a university in Taiwan situated right next to Hsinchu Science Park. Super psyched for this btw. It’s an incredible opportunity to work on this topic with some of the world’s best professors in the field, a street across HQs of billion dollar powerhouses like TSMC and Mediatek, at the epicenter of semi manufacturing. At the same time, I don’t really know much about the field itself, so there’s a lot of learning to be done!

Why sys trading? For one, the skillset required for QR aligns pretty well with what I have “specialized” in/enjoyed so far. My landmark undergrad project is a paper that I coauthored with ITU on modeling low signal broadband network data, and I spent most of my final year internship analyzing time series data. Although I’ve worked on other topics like pure CV and NLP, I always found applying them to different domains more exciting. Second, I think learning automated trading has a high potential reward:effort ratio for me. If done well, it can become a nice second income source (see: Moondev on YT). If it interests me enough, it could even turn into a career as a quant later since I did my undergrad in HK. Last but not least, it would be pretty epic to learn the intricacies of markets and then blow everything on GME calls.

Why start a blog now? I figured it’s a good time to start a project like this. I have gained some degree of maturity in working on technical stuffs from my undergrad and still have the ability to reasonably control my work hours. Not sure if I’ll be able to maintain this post graduation, but who knows what the future holds. Also, it’s a masters degree. Hence the 2 year timeline.

Now that that’s dealt with, here’s another great nugget from Karpathy on twitter:

How to become expert at thing:

  1. iteratively take on concrete projects and accomplish them depth wise, learning “on demand” (ie don’t learn bottom > up breadth wise)
  2. teach/summarize everything you learn in your own words
  3. only compare yourself to younger you, never to others

I know it’s not realistic to expect to become an expert at anything within a couple of months, but I believe that this study plan should work just as well to quickly gain working knowledge of a previously unfamiliar field. So,

for step 1: I plan to draw motivation for project ideas from papers/twitter/original thought (Claude 3.5 Sonnet) and work on them basis research advice I have found/will find on twitter and related textbooks (naturally, all twitter recommendations).

for step 2: this blog officially counts as my effort to execute on step 2. As stated before, I hope that writing out my thoughts while hacking together these projects will help me build concrete subject knowledge. Note that I tend to fixate on small things and end up spending significant time studying topics that may not pertain directly to the project I’m working on at the moment (specific unknown terminology, random flashback to a bookmarked tweet from earlier, etc). In an effort to keep these posts streamlined, most of that stuff will be omitted here and be stored in my personal obsidian notes vault instead.

for step 3: I’ll try to throw in a takeaways/introspection/retrospection section at the end of each post to get some step 3 action going on in here.

…and that’s about it for the intro. I don’t think there’s anything else relevant that I can add here, so see you in the next one!