Impulse Labs founder on surviving 145% tariffs: hardware startups need breathing room, not panic

Summary

Sam Altman — founder of Impulse Labs, which makes what he describes as the highest-performance induction stove on the market — navigated the 145% China tariff period without a crisis, but not without cost. Impulse paid the full 145% on a handful of early production samples and influencer units imported from China before the pause was announced. The bulk of customer orders landed after the rate dropped, and a three-week delay in the battery pack build turned out to be fortunate: had production run on schedule, the company would have faced a decision on whether to pay a very large customs bill to clear the port.

Tariff response

Rather than scrambling to pre-import a year's supply — which he notes would cost a hardware startup a prohibitive amount of capital — Sam focused on making sure Impulse's suppliers are multinational and have manufacturing sites in Vietnam, Mexico, and other Southeast Asian countries likely to receive favorable tariff treatment. He told existing customers the company would absorb even the 145% rate, and says he's holding to that commitment.

His read during the peak of the uncertainty was that a shock clock was running. Shipping time from China to the US gave him a rough calendar: by early May, something would have to move. Apple's early exemption and the noise from other large importers reinforced that belief. The Geneva de-escalation arrived roughly when he expected.

The longer-term question is whether this episode changes strategic planning in a durable way. His honest answer is that it creates breathing room to think carefully about supply chain — not just for the stove, but for other electrified devices Impulse intends to power. Full decoupling from China is not the plan, but having a documented alternative is now table stakes.

BYD as the competitive benchmark

The more pointed part of the conversation is his analysis of BYD. BYD started as a nickel-cadmium battery manufacturer and kept acquiring businesses until it had a car brand, a lithium-ion battery operation, and essentially all the IP needed to build any electrified device — phones, drones, EVs. Sam describes it as the most vertically integrated company imaginable: infotainment, battery cells, assembly, firmware, all built with BYD parts and BYD process. He frames it as the best Lego set in the world, where a customer can commission a complete device and BYD will build it soup to nuts from its own component library.

Tesla and SpaceX are the only meaningful US analogues. The ex-Tesla source he spoke to suggested Elon Musk would build for third parties only if it sat on the critical path to Tesla's vertically integrated goals — not as a general contract manufacturing play. Invoking the Defense Production Act to force Tesla into that role, Sam says, would go over like a lead balloon and isn't a coherent strategy anyway.

Where the US is and isn't competitive

His breakdown of the iPhone supply chain is clarifying. Strip out final assembly and most of the core engineering IP sits with the US and its Asian allies: Foxconn is Taiwanese, the camera module comes from LG Innotek, the display from Samsung. China's actual edge is in assembly, integration, and the ability to have engineers across disciplines talk to each other within a single organization — the opposite of the legacy auto or appliance model, where electronics box A and metal enclosure B are managed across a corporate firewall.

The broader claim, which he credits to economist Noah Smith, is that China has explicitly identified the electric tech stack — batteries, power electronics, motors, rare earth magnets — as the technology platform of the future, and has built toward it systematically. The US largely missed the window after 2008, when the recession and a lack of subsequent manufacturing investment meant the country never developed that capability at scale. Tesla and SpaceX are the exceptions, and Sam's framing is that understanding how Musk built those twice is the right question for anyone thinking seriously about American reindustrialization.