Lenny's PodcastChip Huyen: Why RAG wins come from data prep, not vector DBs
Preparing data and talking to users beats agonizing over which vector database; Huyen says post-training, not new models, drives real AI product wins.
Chip HuyenguestLenny Rachitskyhost
CHAPTERS
- 0:00 – 4:28
Introduction to Chip Huyen
- 4:28 – 7:05
Chip’s viral LinkedIn post
- 7:05 – 8:50
Understanding AI training: pre-training vs. post-training
- 8:50 – 13:55
Language modeling explained
- 13:55 – 15:20
The importance of post-training
- 15:20 – 22:23
Reinforcement learning and human feedback
- 22:23 – 31:55
The importance of evals in AI development
- 31:55 – 38:50
Retrieval augmented generation (RAG) explained
- 38:50 – 43:19
Challenges in AI tool adoption
- 43:19 – 45:20
Challenges in measuring productivity
- 45:20 – 49:10
The three-bucket test
- 49:10 – 55:31
The future of engineering roles
- 55:31 – 57:12
ML Engineers vs. AI engineers
- 57:12 – 1:05:48
Looking forward: the impact of AI
- 1:05:48 – 1:08:23
Model capabilities vs. perceived performance
- 1:08:23 – 1:22:35
Lightning round and final thoughts
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome