Richard Y. Liu

About me

I'm working on hardware superintelligence to empower AI infra and individuals.


I'm a self-taught engineer, with a strong enthusiasm in linguistics, politics, history, philosophy, economics, literature, and geography. Born and raised in Shanghai, I was in the Gaokao system before dropping out of high school to build robots and self-teach everything I love.


I'm in the inaugural cohort of the Neuro One Fellow, a fellowship program for the best builders from China. I'm also a four-time USACO Platinum holder, with a best ranking of #23.


I'm studying Computer Science and Cognitive Science at UC Berkeley.

I'm working on hardware superintelligence to empower AI infra and individuals.


I'm a self-taught engineer, with a strong enthusiasm in linguistics, politics, history, philosophy, economics, literature, and geography. Born and raised in Shanghai, I was in the Gaokao system before dropping out of high school to build robots and self-teach everything I love.


I'm in the inaugural cohort of the Neuro One Fellow, a fellowship program for the best builders from China. I'm also a four-time USACO Platinum holder, with a best ranking of #23.


I'm studying Computer Science and Cognitive Science at UC Berkeley.

Connect with me on LinkedIn, X, and richard [dot] yliu [at] berkeley [dot] edu.

Vision

We believe that:


  • we can build the infrastructure layer for Agents, Humanoids, and AGI, equipping them with deep context and semantic understanding.

  • there are superior alternatives to VLA (Vision-Language-Action) models for achieving high-performance robotic control.

  • the humanoid form factor is not the optimal solution for replacing human labor, and we are developing a more efficient alternative.

  • the maturation of SOTA models enables us to build a generational company and create significant wealth on a compressed timeline.

Writings

Getting Started

Dec 22, 2025

Investing

I invest in startups and labs founded by friends, as well as the broader infrastructure sector.


I believe that a successful business must possess:

  • a defensible moat (e.g., proprietary data, supply chain dominance, or exceptional human capital).

  • a mission to solve acute, existential problems (e.g., fundamental survival needs).

Previous Works

    • FRC Robotics, Team 6353 Zodiac

      • 3-DoF mechanical arm, 180° pivoting prop intake, and F-style climbing hook.

      • Coaxial ground prop intake and rack-and-pinion angle-adjustable shooter.

      • 1-stage elevator and pneumatic intake.