Show: Training Data · Publisher: Sequoia Capital · Host: Sonya Huang, Pat Grady
Episode URL: https://pscrb.fm/rss/p/traffic.megaphone.fm/CPUAI8200472129.mp3?updated=1768421374
Publish date: 2026-01-14
Duration: NAs
Default source credibility: HIGH — Sequoia partners interview frontier-lab founders + F500 AI buyers. VC-hosted — portfolio-company framing on recommendations; named guest metrics stay HIGH. Peer-tier to No Priors in quality.
- Chip design is a bottleneck for advancing AI, with a mismatch between AI method development and chip design cycles.
- Recursive Intelligence uses AI to optimize chip design, enabling faster and more efficient chip creation, potentially transforming industries from fabless to designless.
- The company aims to democratize chip design, allowing any company with a workload to benefit from custom silicon without needing large in-house design teams.
Extracted quotes
| # | Credibility | Speaker | Org | Timestamp | Topic | Quote |
|---|---|---|---|---|---|---|
| 1 | HIGH | Anna Goldie (Co-founder) | Recursive Intelligence | 09:12 | 01-ai-native-landscape | We saw these very strange curved placements. So there are donut shapes as well. I think humans would tend to make the macros very aligned. And then they would put logic maybe in the middle, and then the wires would connect all of these components. But if you make the shapes curved, you can reduce the wire lengths, which can reduce the power consumption and timing violations. |
| 2 | HIGH | Azalia Mirharseni (Co-founder) | Recursive Intelligence | 14:13 | 01-ai-native-landscape | We saw across each successive generation of TPU that we were being adopted in more and more of the chip block and more of the area and also that we were getting increasingly superhuman performance. |
| 3 | HIGH | Azalia Mirharseni (Co-founder) | Recursive Intelligence | 21:12 | 01-ai-native-landscape | We want to enable any customer that has a workload or a family of workloads that they want to run, that they’re running it at sufficient scale and they would benefit from custom silicon. We want to enable them to have that without having teams of hundreds to thousands of chip designers in their companies. |
Per-quote detail
1. Anna Goldie — Recursive Intelligence (09:12)
We saw these very strange curved placements. So there are donut shapes as well. I think humans would tend to make the macros very aligned. And then they would put logic maybe in the middle, and then the wires would connect all of these components. But if you make the shapes curved, you can reduce the wire lengths, which can reduce the power consumption and timing violations.
- Credibility: HIGH — Named exec with specific claim about AI-generated chip design improvements.
- Topic tag:
01-ai-native-landscape
2. Azalia Mirharseni — Recursive Intelligence (14:13)
We saw across each successive generation of TPU that we were being adopted in more and more of the chip block and more of the area and also that we were getting increasingly superhuman performance.
- Credibility: HIGH — Named exec with specific claim about AI performance improvement over multiple generations of TPU.
- Topic tag:
01-ai-native-landscape
3. Azalia Mirharseni — Recursive Intelligence (21:12)
We want to enable any customer that has a workload or a family of workloads that they want to run, that they’re running it at sufficient scale and they would benefit from custom silicon. We want to enable them to have that without having teams of hundreds to thousands of chip designers in their companies.
- Credibility: HIGH — Named exec with specific claim about enabling custom silicon for large-scale workloads.
- Topic tag:
01-ai-native-landscape
Extracted 2026-04-14T17:27:26 via scripts/podcast_mine.py (MLX mlx-community/Qwen2.5-32B-Instruct-4bit).