One-Year-Old AI Lab Flapping Airplanes Raises $1B at $6B Valuation

Flapping Airplanes, a startup barely a year old, has raised a $1 billion Series A round at a reported $6 billion valuation — just nine months after taking a $180 million seed round at a $1.5 billion valuation. That's a quadrupling of the company's paper value in three quarters, co-led by Index and Lux, the company announced Wednesday.

The speed and size of the round put Flapping Airplanes in rare company. Most startups never raise a billion dollars at any stage; this one did it before its first birthday, on the strength of a founding team and a technical thesis rather than shipped revenue.

Who put in the money?

The new round drew both new and returning capital. Earlier backers Sequoia, GV and Kleiner Perkins participated, according to the company. Valor, Atreides, A* and Nvidia are also investors in Flapping Airplanes.

The seed round nine months ago came from GV, Sequoia and Index. Index has now doubled down by co-leading the Series A — a signal that at least one of its original backers liked what it saw enough to write a much bigger check.

Who are the founders?

The company was founded by Aidan Smith and brothers Asher and Ben Spector. Smith is a former founder and Thiel Fellow who worked at the brain implant company Neuralink during college. Asher Spector was a statistics PhD candidate at Stanford. Ben Spector was a computer science PhD student at Stanford and founded the startup accelerator Prod while at MIT.

Prod itself is a notable part of the story. It's a five-year-old, student-run, almost secret-society accelerator that, as Sequoia partner David Cahn — a Flapping Airplanes investor — wrote in a post earlier this year, has "become one of the most powerful talent incubators in Silicon Valley." Cahn went further in describing the team: together, the founders "are building the 'young person's AGI lab.'"

What does the technology actually change?

The premise is efficiency. The founders describe their work as "the world's most complete megakernel framework. It can perform entire training runs, serve inference, play video games, and even do all three simultaneously in a single kernel." A kernel is a program that runs on a GPU; a megakernel bundles many tasks into one program.

The claimed trick is overlapping work. The framework "can overlap memory-bound inference with compute-bound training," the founders say — meaning it can run a task that mostly waits for data to be fetched alongside one that mostly does math, so the chip rarely sits idle.

The bigger claim is about data. The founders argue AI has barely scratched the surface of what GPUs can do, and that intelligent systems can be built "without ingesting half the internet." The goal is systems that cost less to train and run — the core expense line for anyone building or deploying models today.

"A model that's a million times more data efficient is probably a million times easier to put into the economy," Asher Spector told TechCrunch shortly after the startup launched. That's the practical bet: if training and inference get radically cheaper, the economics of the entire AI stack shift.

What happens next?

The company now has a fresh $1 billion to hire and to try to turn data efficiency from a thesis into shipping infrastructure. For practitioners, the question is whether a megakernel approach can actually deliver cheaper training runs at scale — and whether a one-year-old lab can compete with the compute budgets of the industry's giants.