The journey of Ricursive Intelligence's co-founders, Anna Goldie and Azalia Mirhoseini, appears to have been intrinsically linked from the outset.
Goldie, CEO, and Mirhoseini, CTO, are highly esteemed figures within the artificial intelligence community. Their prominence is such that they were among the AI engineers who "got those weird emails from Zuckerberg making crazy offers to us," Goldie lightheartedly shared with TechCrunch, noting they declined the overtures. Their professional synergy began at Google Brain, where they worked together, followed by a stint as early employees at Anthropic.
Their tenure at Google was marked by significant achievement: the creation of the Alpha Chip. This innovative AI tool was capable of generating robust chip layouts in mere hours, a task that traditionally demands human designers a year or more. The Alpha Chip’s impact was substantial, contributing to the design of three generations of Google’s Tensor Processing Units.
This distinguished background undeniably contributed to Ricursive's rapid ascent. Just four months after its launch, the company announced a substantial $300 million Series A funding round last month, valuing the startup at $4 billion. This followed closely on the heels of a $35 million seed round, led by Sequoia, raised only a couple of months prior. The Series A was led by Lightspeed.
Ricursive distinguishes itself by developing AI tools specifically for chip design, rather than manufacturing the chips themselves. This approach sets them apart from most other AI chip startups, positioning them not as a rival to industry giants like Nvidia, but as a crucial enabler. In fact, Nvidia is an investor, and the GPU powerhouse, alongside AMD, Intel, and other chip manufacturers, represents Ricursive's core customer base.
"We want to enable 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 AI to do that," Mirhoseini articulated to TechCrunch, outlining the company's ambitious mission.
Their shared professional trajectory began at Stanford, where Goldie pursued her PhD while Mirhoseini taught computer science. From that point, their careers progressed in remarkable synchronicity. Goldie recounted their parallel path: "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 rejoined Google on the same day, and then we left Google again on the same day. Then we started this company together on the same day."
During their time at Google, their close collaboration extended beyond work, as they even engaged in circuit training workouts together. This shared routine inspired Jeff Dean, the renowned Google engineer and their collaborator, to playfully nickname their Alpha Chip project "chip circuit training." Internally, the duo was affectionately known as A&A.
While the Alpha Chip garnered significant industry attention for its innovation, it also sparked controversy. Wired reported that in 2022, a Google colleague was dismissed after years of attempting to discredit A&A and their chip work, despite its instrumental role in producing some of Google’s most critical, "bet-the-business" AI chips.
Ultimately, their Alpha Chip project at Google Brain served as the foundational proof of concept for Ricursive: leveraging AI to dramatically accelerate the complex process of chip design.
The inherent challenge in chip design lies in precisely placing millions to billions of logic gate components onto a silicon wafer. Human designers typically dedicate a year or more to this intricate task, ensuring optimal performance, power efficiency, and other specific design requirements. The digital determination of such infinitesimally small components with the necessary precision is, by its nature, exceptionally difficult.
The Alpha Chip, as Goldie explained, "could generate a very high-quality layout in, like, six hours. And the cool thing about this approach was that it actually learns from experience."
The core principle of their AI chip design methodology involves employing a "reward signal" that evaluates the quality of a design. The AI agent then utilizes this rating to "update the parameters of its deep neural network to get better," Goldie elaborated. Through the completion of thousands of designs, the agent achieved remarkable proficiency and, as the founders noted, became progressively faster with each learning iteration.
Ricursive’s platform aims to advance this concept further. The AI chip designer currently under development will "learn across different chips," Goldie stated, implying that each chip designed will contribute to making the AI a more adept designer for subsequent projects.
Moreover, Ricursive’s platform incorporates Large Language Models (LLMs) and is designed to manage the entire chip design workflow, from initial component placement through to comprehensive design verification. Any enterprise involved in electronics manufacturing and requiring chips is a potential target customer for their innovative solution.
Should their platform continue to demonstrate its immense potential, Ricursive could play a pivotal role in the ambitious pursuit of artificial general intelligence (AGI). Indeed, their ultimate vision involves AI designing its own computer brains, thereby creating AI chips.
"Chips are the fuel for AI," Goldie emphasized. "I think by building more powerful chips, that’s the best way to advance that frontier."
Mirhoseini further highlighted how the protracted chip-design process currently constrains the pace of AI advancement. She expressed confidence that Ricursive can "enable this fast co-evolution of the models and the chips that basically power them," thereby accelerating the rate at which AI can become smarter.
While the notion of AI autonomously designing its own brains at increasing speeds might evoke dystopian images reminiscent of Skynet or the Terminator, the founders are quick to point to a more immediate, positive, and, they believe, more probable benefit: hardware efficiency.
When AI labs are empowered to design significantly more efficient chips—and eventually all underlying hardware—the growth of AI will require a substantially smaller consumption of the world’s finite resources.
"We could design a computer architecture that’s uniquely suited to that model, and we could achieve almost a 10x improvement in performance per total cost of ownership," Goldie projected, underscoring the tangible benefits.
Although the nascent startup has not yet disclosed its early customers, the founders confirm they have engaged with every major chip-making entity imaginable. Unsurprisingly, they find themselves in an enviable position to selectively choose their initial development partners.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.