The co-founders of startup Recursive Intelligence seemed destined to be co-founders.
Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got these weird emails from Zuckerberg making crazy offers to us,” Goldie told TechCrunch, laughing. (They declined the offers.) The pair worked together at Google Brain and were employees at Anthropic from the start.
They won recognition at Google by creating the Alpha Chip — an AI tool that could create stable chip layouts in hours — a process that typically takes human designers a year or more. The tool helped design three generations of Google’s Tensor Processing Units.
That pedigree explains why, just four months after launching Recursive, last month they announced a $300 million Series A round at a $4 billion valuation led by Lightspeed, just two months after raising $35 million led by Sequoia.
Recursive builds AI tools that design chips, not the chips themselves. This makes them fundamentally different from almost every other AI chip startup: they’re not an out-of-the-box competitor to Nvidia. In fact, Nvidia is an investor. The GPU giant, along with AMD, Intel and every other chip maker, are the startup’s target customers.
“We want to allow any chip, like a custom chip or a more traditional chip, any kind of chip, to be built in an automated and very accelerated way. We’re using artificial intelligence to do that,” Mirhoseini told TechCrunch.
Their paths first crossed at Stanford, where Goldie earned her PhD while Mirhosseini taught computer science classes. Since then, their careers have been gradual. “We started at Google Brain on the same day. We left Google Brain on the same day. We joined Anthropic on the same day. We left Anthropic on the same day. We came back to Google on the same day and then left Google again on the same day. Then we started this company together on the same day,” Goldie recounted.
Techcrunch event
Boston, MA
|
June 23, 2026
During their time at Google, the colleagues were so close that they worked out together, both enjoying circuit training. The pun was not lost on Jeff Dean, the famous Google engineer who was their partner. He called their Alpha Chip project “chip circuit training”—a play on their shared training routine. Internally, the pair also took on a nickname: A&A.
Their Alpha Chip gained industry attention, but also attracted controversy. In 2022, one of their colleagues at Google was fired, Wired reportedafter spending years trying to discredit A&A and their chip work, even though that work was used to help produce some of Google’s most important, betting AI chips.
Their Alpha Chip project at Google Brain proved the concept that would become Recursive — using AI to dramatically speed up chip design.
Chip design is hard
The point is that computer chips have millions to billions of logic gate elements built into their silicon wafer. Human designers can spend a year or more putting these components on the chip to ensure performance, good power usage, and any other design needs. Digitally determining the placement of such infinitesimal components with precision is, as you might expect, difficult.
Alpha Chip “could create a very high-quality layout in, say, six hours. And the nice thing about that approach was that it really learns from experience,” Goldie said.
The principle of their work in AI chip design is to use “a reward signal” that evaluates how good the design is. The agent then takes that score to “update the parameters of its deep neural network to get better,” Goldie said. After completing thousands of plans, the agent became very good. It also got faster as it learned, the founders say.
Recursive’s platform will take the idea further. The AI chip designer they’re building will “learn on different chips,” Goldie said. So each chip it designs will help it become a better designer for each subsequent chip.
Recursive’s platform also uses LLM and will handle everything from component placement to design verification. Any company that makes electronics and needs chips is their target customer.
If their platform proves itself, as it seems likely to do, Recursive could play a role in the goal of achieving artificial general intelligence (AGI). Indeed, their ultimate vision is to design AI chips, meaning that AI will essentially design its own computer brains.
“Chips are the fuel for artificial intelligence,” Goldie said. “I think that by creating more powerful chips, that’s the best way to advance that frontier.”
Mirhoseini adds that the lengthy chip design process limits how fast AI can advance. “We think we can also enable this rapid co-evolution of the models and the chips that basically power them,” he said. So AI can get smarter faster.
If the thought of AI designing its own brains at ever-increasing speeds brings to mind visions of Skynet and the Terminator, the founders point out that there is a more positive, immediate and, they believe, more likely benefit: hardware efficiency.
When AI Labs can design much more efficient chips (and eventually all the underlying hardware), their development won’t need to consume so much of the world’s resources.
“We could design a computer architecture that uniquely fits this model, and we could achieve almost a 10x improvement in performance per total cost of ownership,” Goldie said.
While the young startup won’t name its first customers, the founders say they’ve heard from every big name brand you can imagine. Unsurprisingly, they also have their pick of their first development partners.
