There’s a quote I keep coming back to — attributed to Richard Feynman, though who really knows — that goes something like: “What I cannot create, I do not understand.” I’ve always taken this to mean that the act of building something (a model, an argument, a piece of writing) forces a kind of understanding that passive reading never achieves.
This blog is my attempt at taking that seriously during graduate school.
What This Space Is For
I’m starting an MS + MBA, and the next two years will involve a density of new ideas that I won’t be able to absorb without some system for processing them. This blog is that system. Expect a mix of:
- Paper reviews — distilling academic papers into their core ideas, with commentary on what’s interesting and what’s missing
- Code experiments — when I implement something from a paper or class, I’ll write it up here with code and results
- Research dispatches — notes from ongoing research, including dead ends
- Reflections — on the experience of grad school, career thinking, and the intersection of technology and business
A note on audience: I'm writing primarily for myself — as a learning tool. But I'm keeping it public because I believe in working with the garage door open. If you find something useful here, that's a bonus.
The Broad Knowledge Base Philosophy
Throughout my career, I’ve optimized for breadth over depth. At dive.ai, I was building LLM chatbots one week and redesigning cloud infrastructure the next. Before that, I was doing econometrics for a think tank while teaching linear algebra at a university.
This isn’t accidental — it’s a deliberate strategy. The most interesting problems I’ve encountered sit at the intersection of disciplines. Fraud detection requires understanding both the ML models and the business processes they protect. A competitiveness index requires PCA and political context.
Graduate school is an opportunity to add more tools to this toolkit, and this blog is where I’ll document what I’m learning.
On Format
I’m drawn to the Distill.pub style of academic writing — clear prose, inline data, and a respect for the reader’s time. Not every post will be a polished artifact. Some will be rough notes. But I’ll try to make each one worth reading.
The plan is one post per week, though I suspect the reality will be more irregular. Quality over quantity.
What’s Next
The first technical posts will likely cover material from my early coursework — probably something in optimization or statistical learning theory. I’m also planning a series of paper reviews on topics I want to go deeper on before my research gets underway.
If you want to follow along, there’s an RSS feed — the most civilized way to follow a blog.
Let’s see where this goes.