The Growth Project
From 17% to 97.8%: Making a Laptop-Sized AI Actually Reliable
About this episode
A small on-device model started at 17% on a real task and ended at 97.8% with full speed intact. The biggest lever wasn't the model. It was the prompt.
Sam and Maya unpack From 17% to 97.8%: Making a Laptop-Sized AI Actually Reliable from the field notes, the argument, the numbers, and the practical moves, in a short, sharp conversation.
Chapters
- 0:03 From 17 to 97.8
- 0:52 Lean prompts for small models
- 2:22 Taxonomy, not model, problem
- 3:42 Full accuracy, full speed
- 4:57 Small models can do real work