I'm a researcher, historian, and investor.
I am currently working on post-training at xAI.
Previously, I was a student at UC Berkeley, where I studied computer science, applied math, and business. I plan on re-enrolling eventually, graduate, and pursue a PhD in computer science.
During my time as a student, I interned at HOF Capital, where I built proprietary software for the firm and led several investments.
In high school, I competed in the USA Computing Olympiad and held a fond interest for history, especially that of Julius Caesar and Napoleon.
More to come.
My current interests revolve around hardware-software co-design, reasoning, and multi-agent systems.
I strongly believe demand for compute will increase exponentially over time, and I believe general intelligence for most tasks will be commoditized in the form of open-weight models and will be hosted not on personal devices, but large superclusters. To be clear, I do believe that closed frontier models will continue to demand a premium for their marginal improvement in intelligence for the most pressing tasks.
Following on my belief in the commoditization of general intelligence, I believe personal taste and intuition will matter significantly more in the coming future. This also relates to my conviction in personal AI, in which general models are too excessively fine-tuned to work for everyone's specific tasks and contexts.
I will write checks to friends and into startups going after frontiers I'm passionate about, especially those working in continual learning, infrastructure and inference for AI (energy, data, hardware), and reshoring US manufacturing.
A space I'm particularly interested in is infrastructure and technologies that enable robotics and autonomous machines to more efficiently coexist in a fundamentally human world.