The AI Report Podcast

How Your Voice Reveals Emotion, Deception, and Intent | Rana Gujral, former CEO, Behavioral Signals

14 snips
Jul 30, 2026
Rana Gujral, former CEO of Behavioral Signals and author of The AI Instinct, pioneers voice-based behavioral AI. He explores how 75–100 vocal dimensions reveal intent, stress, deception, and physiology. Conversations on deepfake detection, call-center matching, pitch compression as a stress signal, ethics of emotion AI, and his Artificial General Experience idea make for a provocative listen.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
INSIGHT

What Makes Conversations Click

  • Compatibility between two speakers shows as entrainment: synced pitch contours, converging speaking rates, aligned breath and mirrored energy.
  • Incompatibility appears as asymmetric turn-taking, longer latencies, missing prosodic mirroring, and neurologically felt effort.
INSIGHT

Voice As A Remote Physiological Sensor

  • The human voice already encodes physiological signals like heart rate variability, respiration, and muscle tension via microtremors, phrase length and jitter.
  • Rana treats voice as a non-invasive multi-physiological sensor usable over phone, Zoom, or car speakers.
ANECDOTE

How Studying Voice Changed Rana's Speaking

  • Studying voice changed Rana: he now leaves longer pauses, moderates pitch and listens on a 'second channel' of jitter and micropauses.
  • That awareness improved empathy but also made his interactions slightly more performed and deliberate.
Get the Snipd Podcast app to discover more snips from this episode
Get the app