Ricursive Intelligence is betting on AI-designed chips to help solve the compute crisis

Summary

Ricursive Intelligence, founded by Anna Goldie and Azalia Mirhoseini, presented a system at TechCrunch Disrupt 2026 that performs end-to-end chip design using artificial intelligence. The company claims it can reduce design cycles from months to weeks and generate unconventional layout patterns, such as 'bagel' shapes, that outperform traditional block-based approaches.

Why it matters

If computing demand truly increases by 10,000 times within five years, the bottleneck will not only be capital for data centers but also the speed and efficiency of silicon design. A functional AI-assisted chip design cycle could democratize custom silicon in niche markets that are currently economically unviable (AR/VR, space, bio-computing). This could reshape the fabless semiconductor value chain.

Highlights

• AlphaChip's circular 'bagel' layout initially surprised engineers but demonstrated the validity of the approach.
• Unlike traditional block-based workflows, the team optimizes the entire chip layout, enabling wave-like and other non-standard patterns.
• Goldie positions the human role in idea generation and the AI role in execution, describing chip design as the next field to follow software in adopting this distinction.

Critical take

The article reads like a verbatim summary of a conference presentation. It lacks independent benchmarks, named customers, or data on efficiency and power consumption. The '10,000x demand' headline is presented as the founder's own 'conservative' estimate. The claim of 'chip design in a week' is made without specifying any conditions. Since complexity, manufacturing node, and verification scope are never explained, the text reads more like a promotional summary than a reportable development.