xAI co-founder's River AI raises $1B for home AI computers
A year after leaving the lab he co-founded with Elon Musk, Igor Babuschkin is back with a billion-dollar war chest and a contrarian thesis: the next big AI platform shouldn't live in a data center. It should live in your house.
xAI co-founder Igor Babuschkin's new startup River AI has raised $1 billion — a round led by General Catalyst, per a Techmeme report on a New York Times story — to build computers for homes and small businesses that run AI locally, owned and controlled by the user rather than any lab. The company's pitch, as the Times describes it, is "a new kind of computer inside people's homes" that people can own, modify and retrain without oversight from anyone else — including River AI itself. The startup plans to release its AI technologies as open-source software, free for businesses, independent developers and consumers to use and alter. "We hope that in the future, A.I. will be trained to benefit you as the individual, rather than being a centralized thing that is controlled by one big corporation," Babuschkin told the Times.
The round caps a fast trajectory for a company that was only a rumor in the spring: Forbes reported in May that Babuschkin was in talks to raise up to $1 billion at a valuation of up to $5 billion, with General Catalyst in discussions to lead. Babuschkin co-founded xAI with Musk in 2023, helped stand up its Memphis supercomputer in roughly four months, and left in August 2025; he now runs River AI as co-founder and CEO.
The bet here is that the pendulum swings back from centralized clouds. River AI is the highest-profile attempt yet to make "personal AI" literal: a box in your home holding your models, your data, your fine-tunes — with open weights as the default so the company has to win on capability rather than lock-in. The hard questions are practical ones. A home server will struggle to keep pace with frontier labs that spend billions per training run, and most consumers have shown little appetite for operating their own AI stack. But General Catalyst's answer to both objections is a $1 billion check: it's worth finding out whether local, user-owned AI can carve out a real third way between cloud oligopoly and hobbyist tinkering.
What to watch: whether River AI ships actual hardware or starts with software, and how "trainable" that home box really is.
Would you put a trainable AI server in your house, or is cloud AI good enough? Tell us in the comments.
