Ricursive Intelligence to outline AI-driven chip design at Disrupt
Ricursive Intelligence co-founders Anna Goldie and Azalia Mirhoseini will appear at TechCrunch Disrupt 2026 to discuss AI systems intended to automate and accelerate chip design. The startup says it wants to reduce a process that can take two to three years to a cycle measured in weeks.
The session, titled When AI Starts Designing Its Own Hardware, will examine a feedback loop in which AI helps create the hardware used for more capable AI systems. Goldie and Mirhoseini are scheduled to explain the technical bottlenecks and the company’s goal of building systems that improve through repeated design work.
From AlphaChip to Ricursive
Before founding Ricursive Intelligence in late 2025, Goldie and Mirhoseini co-led AlphaChip at Google. The AI system could generate chip layouts in hours rather than through the substantially longer process used by human designers, and their work helped design multiple generations of Google Tensor Processing Units.
Ricursive is extending that premise with tools intended to automate more of the chip-design workflow, including component placement and design verification. Its system is designed to learn across different chips, allowing work on one design to inform its approach to the next.
The founders also co-founded Google’s ML for Systems team. Both were early employees at Anthropic and senior staff research scientists at Google DeepMind. Goldie is Ricursive’s founder and CEO, while Mirhoseini is its founder and CTO, an assistant professor of computer science at Stanford, and founder of the university’s Scaling Intelligence Lab.
Capital and the hardware-development gap
Within four months of launch, Ricursive raised $335 million at a $4 billion valuation, including a $300 million Series A. Nvidia is among the company’s investors. The funding reflects the scale of the opportunity Ricursive sees in making chip development faster while AI capabilities continue to advance.
Hardware has a different development cadence from AI models. Ricursive is betting that AI-assisted design can narrow that gap by carrying learning from one chip programme into later projects. The wider event programme also includes TechCrunch Disrupt sessions on scaling AI on scaling AI, placing chip-design speed alongside the infrastructure questions facing AI builders.
What the founders will address
Goldie and Mirhoseini will discuss self-improving systems and the prospect of a loop in which AI helps design better hardware, that hardware supports more powerful AI, and later systems assist with subsequent designs. Their presentation is one of more than 200 sessions across six stages, roundtables and breakouts at Disrupt, which runs October 13–15 at Moscone West in San Francisco.
For businesses planning AI infrastructure or products dependent on specialised compute, the practical implication is to watch whether automation can reliably shorten both design and verification work, because chip-development speed may become an important constraint on future AI deployment.