A one-year-old AI lab with no product on the market has raised $1 billion to argue that the industry is training its models the wasteful way. Flapping Airplanes announced on Wednesday that it had raised a $1 billion Series A round, co-led by Index and Lux, reportedly at a $6 billion valuation, according to TechCrunch.
The round lands just nine months after the company launched with a $180 million seed round from GV, Sequoia and Index. That seed round had valued Flapping Airplanes at $1.5 billion, so the new valuation has quadrupled in well under a year. Other participants in the Series A included earlier backers Sequoia, GV and Kleiner Perkins, and TechCrunch reported that Valor, Atreides, A* and Nvidia are also investors in the company.
The young person’s AGI lab
Flapping Airplanes 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 at Stanford, while Ben Spector was a computer science PhD student at Stanford and founded Prod while at MIT.
Prod is a five-year-old, student-run startup accelerator that has “become one of the most powerful talent incubators in Silicon Valley,” Sequoia partner David Cahn, an investor in the company, wrote in a post earlier this year explaining why he invested. Together, Cahn wrote, the founders are building “the young person’s AGI lab.”
One kernel, three jobs at once
The technical claim at the center of the company is about how hard graphics processors actually work. 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, and a megakernel bundles many tasks into one program.
The founders say the framework “can overlap memory-bound inference with compute-bound training” — pairing a task that spends most of its time waiting for data to be fetched with one that spends most of its time doing math, so the chip is rarely sitting idle. Their premise is that AI has barely scratched the surface of what GPUs can do, and that better use of the hardware can create intelligent systems “without ingesting half the internet.”
That second goal — data efficiency — is the company’s founding mission: finding ways to train large models with less data. “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.
What $1 billion has to prove
Analysis: The round prices a thesis, not a business. Nothing in the announcement describes a product for sale, and the megakernel and data-efficiency claims are, at this stage, the founders’ own description of their framework. What the $6 billion valuation buys is time and talent: TechCrunch reported the fresh capital gives the company room to hire and attempt to turn the goal into reality.
The bet runs against the industry’s dominant direction. Frontier labs have mostly competed by scaling data and compute; Flapping Airplanes is raising at a frontier-lab scale to argue the opposite — that cost structures can be reshaped by using existing chips and smaller datasets far more efficiently. If the framework delivers even a fraction of that, the economics of training change for every lab that rents GPUs. If it does not, the Series A will be remembered as the peak of investor appetite for research-first labs. Either way, a year-old company now has $1 billion to find out.
