TechCrunch Disrupt 2026 will feature Ricursive Intelligence co-founders Anna Goldie and Azalia Mirhoseini in a session exploring one of AI’s most exciting frontiers: when AI begins helping design the hardware it runs on.
The discussion will focus on closing the loop between AI and chip development. As AI systems become more capable, they may increasingly assist engineers in creating chips that are faster, more efficient, and better suited to demanding AI workloads.
Why this matters
Chip design is a foundational part of AI progress. If AI can help improve the hardware pipeline, it could accelerate innovation across the entire ecosystem—from research labs and cloud platforms to real-world products that rely on powerful computation.
- Smarter tools: AI could support more efficient chip design workflows.
- Better infrastructure: Optimized hardware can help future AI systems run faster and more sustainably.
- Positive feedback loop: Improved chips can enable stronger AI, which in turn may help design even better chips.
While this is a conference announcement rather than a deployed breakthrough, it points to a high-potential direction for the industry: AI helping build the next generation of AI infrastructure.