Jusung

instagram linkedin

About Me:

• I like tech and neuroscience. I like to build things that are interesting to me. Lately that has been mostly around applied AI for my startup.
• I graduated from Dartmouth College in 2025. I majored in cs+neuro and minored in English.
• I was born in South Korea but then moved to North Carolina when I was 5.
• I currently run Centi, a VC-backed company that does process mining and automations for private equity portfolio companies. I also co-founded and am the technical partner of Antidote, a full-stack AI transformation company that operates via buyouts and various forward deployed engagements. We buy companies in the $2-5m EBITDA range (amongst other criteria).
• I'm currently in New York City.
• I took the LSAT during a mid-college career crisis and got a 177.

Some things I believe:

• sublation is one of my favorite concepts.
  • I like Hegel. he coined the word Aufhebung which has three related meanings at once: 1) to cancel or abolish: the lower stage or concept ceases to exist in its original independent form. 2) to preserve or maintain: the essential truths or features of that lower stage are kept intact. 3) to transcend or lift up: it is elevated to a higher, more comprehensive level of truth or reality.
  • in short, my practical application of the term is that given two usually mutually exclusive choices, there is always an other, third option where you can do the best parts of both.
• time dilation is offset by being open to new experiences
  • we relatively experience half of our life by age 9 and 3/4th by age 27 (assuming we live to 80). Very depressing. I've found that doing new things, breaking monotonous cycles, creating new neuronal pathways, have helped stretches of time feel longer, more fruitful, and help maintain childlike wonder.
• binary thinking is stupid. and overall very detrimental for society.
• metacognition is the single greatest skill one can master.
  • video games, even open world sandboxes, make it very easy for the player to progress through relatively quick feedback loops and clear evals. much of life is the same with the time horizons and number of variables an order of magnitude greater and largely unbounded. Having awareness of your own thinking, actions, and outcomes and reflecting+iterating on those can take you very far.
• grind culture is a lie.
  • i am all for working hard for long periods of time. but there is a point on the bell curve past which working becomes motion without direction. working too much is counterintuitively unproductive. plus it's just not fun. you lose the passion you had for the work in the first place and being miserable, stuck hating the things you once loved is not what life is about.
  • it is very easy to overdo the leisure however.
• sustained, deep attention is the most scarce resource in the world today.
  • nothing worthwhile was built on a whim with little care. even with the smartest people in the world, their greatest work is the result of countless hours of painstaking focus and effort. no amount of IQ can offset a lack of continuous, arduous attention to a task at hand.
• know when to listen to others and know when to not
  • a key motif to getting far in life is to not listen to others or what society tells you. this is often true, as it's one of the few ways to find alpha in life. however, this does not apply to every aspect. it's also important to know what you don't know and fill in gaps or have others much further along in some path you want to take help expedite your journey.
  • in addition, it's important to understand the underlying reasons behind decisions and status quos instead of blindly accepting them to help speed up knowledge and skill acquisition. the whys are often much more insightful into an industry or finding out why something ticks a certain tempo than the whats.
• making smart decisions quickly is a superpower.
  • this is highly contingent on one's ability to anchor decisions on educated and fundamental truths (first principles!) and not doing stupid shit for the sake of doing anything. no one truly really knows what they are doing to a great extent, but the fastest way to get information is to act on uncertainty (ideally a path with a two way door instead of an irreversible decision) and the world gives feedback and information that puts you a step ahead. rinse and repeat countless times and you become an expert in those microinteractions between you and the world. much like how when you're playing a game against someone new, it's a black box until you start playing and get the sense of how they react to certain aggression, plays, and tactics.
  • The fastest way to figure out what to do in a field full of question marks is to do something, then it becomes more obvious what is lacking from that first iteration. And then do that over and over.
• i am a huge fan of math and am very artificial superintelligence pilled.
  • i believe you can solve anything with math, and everything we do and understand in any domain is at some abstraction on top of math. even things like intuition or domain expertise can be reduced to pattern recognition and abstracted components of some math fundamental. of course, this has been unnecessarily inconvenient and inefficient way to learn other fields. however, frontier AI becoming exponentially stronger at much of these basic fundamentals will likely give it the capabilities to near-brute force its way via leveraging computation over general methods (see "The Bitter Lesson" by Richard Sutton). It is becoming less and less important for domain-specific information in acting as a moat against industry disruption, as long as the right principles are identified and the domain specific nuances are understood (tbf very hard to recognize all of them from the outside-in.)
• udon is the best type of noodles