US Robotics Is Set to Boom—First, Valves and Switches Need to Get Cheaper

On October 29, 2026, a16z released the American Dynamism 50 list under the theme Production is Deterrence—literally, production is deterrence. The list covers the industrial base end to end, from inputs in the ground to factories, and then to the systems that deter conflict. a16z's Erin Price-Wright, Connor Love, and Elena Burger broke down the list on the podcast. Connor Love's assessment is very specific: the subsystems used in defense systems and future robotic systems use the same things—actuators, motors, valves, and switches.

His logic is not complicated. To build many systems, you need many subsystems underneath. Today, some very interesting military systems are being built. Building these systems will also build capability for another industry along the way. Once these actuators, motors, valves, and switches can achieve both quality and cost, markets that previously did not exist will begin to exist. The US may thereby catch a robotics wave.

AD50 is a16z's annual list, selecting 50 companies building in the direction of America's future, including both a16z portfolio companies and outside companies. This year's theme revolves around one sentence: without an industrial base, there is no defense industrial base. The podcast chapters run from mines to factories and systems, discussing how DJI grew from drones into a robotics giant, the scaling experience of SpaceX and Tesla, Mariana using software to run a copper mine, why AI still cannot turn a wrench, how defense components can unlock American robotics, the rise of the manufacturing middle layer, people coming out of Anduril and SpaceX, and the US capital market's advantage relative to China.

The full podcast is on YouTube, titled 《How to Build the Next SpaceX or Tesla》, link: https://youtu.be/cRJ7CfaK1Vc

Components Need Buyers First

Connor Love mentions a step that is often skipped. Components like actuators and valves sound boring, but without a first buyer, suppliers will not open production lines, build tooling, or drive down costs for them. In the replies, Winston B. says actuators and valves are boring components; without a first buyer, no one opens a production line. If defense is that buyer, robotics startups can inherit the supply chain and save the cost of building one. Kevin Pham's reply is even more direct: first give the valve and actuator suppliers the orders, and the rest of the stack will follow them.

Robotics Doesn't Depend Only on Models

AI still cannot turn a wrench. Someday it can, but that will depend on actuators, switches, and subsystems that have already been built. Connor Love says that once these capabilities are built in the US, there will be no need to rely on Chinese factories. The focus here is the physical layer. In the replies, Arjun Kakarlapudi says everyone talks about AI models, but no one talks about how models are made, who owns the components, and who will run the robotics era. That is a bit absolute, but it points to an overlooked layer.

The Manufacturing Middle Layer

In the replies, gianluca says the manufacturing middle layer is worth watching: everyone funds drones, no one funds foundries. Another user, JianYuan Deng, worries that the bottleneck will be people willing to work on the factory floor. Another reply asks whether defense suppliers can lower unit costs fast enough for commercial robots to afford them. These questions are more critical than the name of the list. Defense orders can provide first-purchase demand, but whether quality, cost, and capacity can all be ramped up at the same time determines whether the robotics wave can land in the US.

What the US lacks may not be another model. What it lacks is the industrial base to make valves, actuators, motors, and switches at the required quality and cost. This a16z list lays out where the money is flowing. What to watch next is whether these component suppliers can first get enough buyers and then bring costs down.