Flapping Airplanes is a San Francisco-based foundational AI research lab founded in January 2026 by brothers Ben Spector and Asher Spector, and Aidan Smith (a former Neuralink researcher and Thiel Fellow). The company's core bet is that today's dominant approach — training ever-larger models on ever-larger datasets — is unsustainable and ultimately a dead end. Instead, it pursues biologically inspired training methods that drastically reduce how much data an AI needs to learn effectively, mimicking how the human brain acquires knowledge from far fewer examples. The name is a deliberate metaphor: where incumbents build "fixed-wing" Boeing-787-sized models requiring massive compute runways, Flapping Airplanes aims to achieve powerful AI through more efficient, adaptive "flapping" approaches. The company operates as a pure research lab with no near-term product or revenue obligations, funded by top-tier VCs willing to bet on long-horizon scientific breakthroughs. Buyers of this research output would ultimately be model developers, cloud providers, and enterprises needing cheaper, more generalizable AI.