Show HN: Sparrow-1 – Audio-native model for human-level turn-taking without ASR

4 points | by code_brian 8 hours ago