Lim Jian Yang Computer scientist and investor MSCS student at Columbia University, New York
I write about AI, investing, life, and random thoughts.
This site is a public journal for working through ideas in the open. Everything here reflects my current thinking, written with the full awareness that there is far more I do not yet understand, and plenty that time will prove wrong.
Projects
- AikenDewit A focused planner for days with too much on your mind. Web app · aikendewit.com · Film · 0:58
- Fencing Scores Score fencing bouts and run competitions. iOS app · App Store
- Value Investing Tools A manual valuation workbench for investors. iOS app · App Store · Film · 1:01
- Build a Thesis An AI research workspace for investors. Web app · Sunset
Over time
86 pieces since August 2024
21 Nov 2026 · Computing · Short idea The Limits of Verifiable Work
Contents
- Investing Longview Investing 17 essays
- Musings The Illusion of Permanence 14 essays
- Computing AI Harnesses and Human Hierarchies 8 essays
- Books Unplugging from Manufactured Values: Nietzsche, Nguyen, and Simone Weil 5 essays
- Cooking Hunan Pork Belly Stir Fry 7 essays
- Short Ideas The Limits of Verifiable Work 35 notes
- Library 33 reads, 15 anime
- Heroes 5 people
- Commonplace Quotations
Recent
-
The Limits of Verifiable Work
Frontier models will master any task with a clear pass/fail check, leaving human value in the messy, unverifiable open world.
Computing · Short idea
-
The Curse of Competence
Jim Collins on why doing what you are merely good at keeps you from becoming great, from Good to Great.
Musings · Short idea
-
We Live in Scary Times
Between rumoured breakthroughs in recursive self-improvement and models already capable of replacing entire teams, progress isn't plateauing.
Computing · Short idea
-
AI Harnesses and Human Hierarchies
Why AI harnesses and multi-agent architectures keep borrowing from human workplace hierarchies, especially given context window and cost limits.
Computing
-
Market Research vs Moving Fast
As iteration gets faster and cheaper, market research meant to predict reactions might become too slow relative to just shipping and seeing what happens.
Musings · Short idea
-
Longview Investing
An experiment in teaching a frontier model to perform true investment analysis on various companies.
Investing
-
Dream big, fellas
A proof by induction that any goal is achievable via an LLM from zero starting knowledge, even when you do not know what questions to ask.
Computing
-
Not Worrying About Office Politics
It's the people who go to work not worrying about office politics that change the world.
Musings · Short idea