a16z tells Fei-Fei Li's story in greentext: 14M labeled images bet that started deep learning
lukas_m_ziegler · x · 2026-09-06
a16z posted Fei-Fei Li's origin story in 4chan-style greentext, and it's spreading: born in Beijing, raised in Chengdu, moved to New Jersey at 16 speaking almost no English, worked the counter at her parents' dry cleaner, then Princeton on a full scholarship (physics) and a Caltech PhD. While everyone bet on algorithms, she bet on data — hand-labeling 14 million images (ImageNet) despite "this will never scale" skepticism. In 2012 a Toronto team trained a convnet on it, error rates fell off a cliff, and the deep learning era began. Then Stanford professor, Google Cloud chief scientist, founder of Stanford HAI.
More from Companies & People
- 82% of enterprises run AI pilots, but over half never deploy: production, not models, is the bottleneck — worldofjaved · 2026-09-06
- GiveWell Scales Research Rigor With On-the-Ground Data, Red Teaming and AI Tools — RexDouglass · 2026-09-06
- Martin Casado floats GMI — General Machine Intelligence — as a better term — viksit · 2026-09-06
- a16z's Martin Casado declares MGI era has arrived, praises OpenAI and Anthropic — viksit · 2026-09-06
- Domingos: Jensen distrusts OpenAI and Anthropic, so Nvidia builds its own models — pmddomingos · 2026-09-06
- ChatGPT Ads rolls out to select MENA countries, with UAE and Saudi Arabia notably absent — yassermasood · 2026-09-06