Stephen Brouhard(@ssbrouhard):Awesome innovation from the @trycua team! We dont need to burn big model inference just to decide what element of a UI to interact with. Using a local small specialized model to handle execution is the way. Cua-S1 is applying almost the same philosophy as Jev, one layer further down. • screen + candidate actions ↓ S1 ↓ • element • action • abstain

Stephen Brouhard

@ssbrouhard

Agent tools that verify · PackSift, AXI CLIs, npm release hardening · quality gates over slop

加入 April 2019

504 正在關注 754 粉絲

Stephen Brouhard@ssbrouhard

2026.09.23 19:49

Awesome innovation from the @trycua team! We dont need to burn big model inference just to decide what element of a UI to interact with. Using a local small specialized model to handle execution is the way. Cua-S1 is applying almost the same philosophy as Jev, one layer further down. • screen + candidate actions ↓ S1 ↓ • element • action • abstain