Andrew Kang

A beautiful portfolio

Dyna just dropped a world model for robotics achieving a new record for scale of pretraining using human video data - 1 million hours This is 2 orders of magnitude larger than NVIDIA's Egoscale dataset (~20k hours) They established scaling laws parallel to the scaling laws

Dyna just dropped a world model for robotics achieving a new record for scale of pretraining using human video data - 1 million hours This is 2 orders of magnitude larger than NVIDIA's Egoscale dataset (~20k hours) They established scaling laws parallel to the scaling laws identified in GPT-3. Pretraining on more human video data increased robot performance and generalization across 4 orders of magnitude Most importantly, the scaling fit power laws with no plateau

This is the most exciting moment of my life so far. And it’s only the beginning. Scalability is EVERYTHING. Over the next few weeks, we’re going to unpack 6 breakthroughs that we believe are critical to scaling physical intelligence: Foundation models ← this is the first big

Today we are introducing Dyna-2, a world-action model pre-trained on one million hours of human video. At this scale, for the first time, we discovered several new scaling laws: • world-action models exhibit scaling law on human data across four orders of magnitude, from 1000 t.co/wZamR0axzS

A new breakthrough is coming. Stay tuned. x.com/DynaRobotics/status/2085…

Portfolio company @PathRobotics just signed a $600M+ agreement with Huntington Ingalls to accelerate naval ship production. Big step for Path Robotics and a win for the robotics industry overall. t.co/DkF4bKzc7e

MAD Society | S1 Physical AI [2h Speed Round] x.com/i/broadcasts/1yKAPPjmnpa…

We get asked a lot when will Physical AI systems start getting deployed and start generating real revenue The answer is it’s already happening. @PathRobotics just signed a $600m contract with one of the largest shipbuilders in the world x.com/Rewkang/status/208539554…

The notion that all hardware is a commodity business is going to be flipped on its head this decade Unitree's IPO filings show that it has ~60% gross margins, higher than most AI labs Durable, high quality robots are going to be supply constrained. Developer and skill ecosystems will create moats around certain robot platforms. Companies that can create an amazing product will command brand premiums The force of capitalism is expanding beyond bits scmp.com/tech/tech-trends/arti…

It’s always surprising to me that a chunk of the investment community is still skeptical of Humanoids and Figure The same was true for reusable rockets and SpaceX until more recently But this is what creates great investment opportunities. Fun interview with Jose.

The former US National Security Advisor is advocating for a National Strategic Investment Fund with an initial capitalization of $50 Billion I tend to agree x.com/Rewkang/status/208303552…

This is how Gemini Robotics 2 helps @Apptronik’s Apollo 2 use whole body intelligence to pack for a sports game ↓ t.co/rPYsclzJHR

Bringing private robotics exposure to the public markets x.com/RoboStrategy/status/2082…

Invest in Robotics x.com/RoboStrategy/status/2082…

Nice to see our portfolio companies @Apptronik and @DexmateAI supporting research at Google

One brain. For any robot. 🤖 We’re launching Gemini Robotics 2: our next-generation physical AI bringing full body intelligence to humanoids, advanced dexterity, multi-robot teamwork and more. t.co/1rFw3NUytb

The US just banned Chinese robots. In our interview from last month, Andrew Kang mapped out the exact playbook - EV tariffs first, robots next. "We want to build things in America, and to do that, we can't be relying on robots from other countries." "I'm not saying I t.co/elIiPOWPIv

Deep dive on how industrial robots are designed and how that design is evolved for an AI native world

America invented the robot Production then moved overseas Standard Bots is the only company designing and producing industrial robots at scale in the US

Standard Bots - A Short Film An inside look at @StandardBots and the team helping shape the future of the US robotics industry t.co/2bI5MpXL4z

In 2009 we had a too big to fail tailwind for established companies In this era of government, we have a too important to not succeed mandate for smaller companies Golden era for venture capital and entrepreneurs in critical industries

The @FCC 's Title III equipment certification power under the Communications Act gives them authority over devices that emit or use radio frequency—WiFi, cellular, proprietary wireless comms, spectrum-based sensors. If a humanoid robot contains any wireless capability, it falls

"FCC plans to announce restrictions Tuesday barring imports of new Chinese humanoid and quadruped robot models, along with new Chinese power inverters, according to U.S. officials" If it wasn't clear enough, the USG is putting its full weight behind the domestic robot industry

The best favor you can do for yourself is to learn about the history of robots, how they are designed, and how that design is evolving to enable these machines to be intelligent x.com/RoboStrategy/status/2081…