/Brian Kihoon Lee

LLM Shibboleths Determine AI Effectiveness tl;dr: “Coding assistants promise to revolutionize software development, but why do some developers sing praises while others find them useless? The answer lies between the keyboard and the chair, but it’s more than just simple user error. Your level of expertise silently shapes the way you interact with the AI, allowing two people to have completely different experiences despite interacting with the same AI on the same subject. In this essay I’ll discuss how this is possible and what you can do about it.”

featured in #620


LLM Ecosystem Predictions tl;dr: "I don’t think I’m making any outlandish claims here - merely simple extrapolations of trends that I think have robust foundations, with a dash of history. And yet the conclusions are surprising. The world of 2030 will be as unrecognizable to us as the world of cellphones today was in 2010, and as the world of the connected Internet was in 2000."

featured in #594


Who Pays You? And Why? tl;dr: “I get asked for career advice from time to time. While each situation is different, a recurring theme is disempowerment - feeling like there’s nothing you can do to advance your career. To help diagnose, I like to ask two questions: Who pays you? And why? These two questions encourage you to leave the comfort zone of job descriptions and confront the reality of what it’ll take to get to the next level at your current job, or potentially a new job. I’ll explain how I think about these questions, and this hopefully helps you think through your own situation.”

featured in #512


Simplifying Fluffy Constructors In Unit Tests tl;dr: Brian discusses the challenges of writing unit tests that become bloated with unnecessary details. “A very common problem is that, over time, objects accumulate fields and subobjects, until it takes significant effort just to construct an object.” To address this, he proposes two solutions: (1) Factory methods: hide irrelevant details, making it easier to write and read tests. (2) Domain-Specific Languages: reduce syntactic fluff, making the code more readable and maintainable.

featured in #451


Readability: Google's Temple To Engineering Excellence tl;dr: Brian shares his experience as a readability mentor at Google and reflects on its cultural significance within the company. While he doesn't recommend implementing Google's version of readability in other companies, he proposes a variant called "Readability Lite" that focuses on consensus on readability standards, mentorship programs, and non-blocking mechanisms to encourage engineers to strive for mastery.

featured in #430