Cross-talk is portrayed as a problem to AIs here, but as someone with Auditory Processing Disorder, I would pay serious money for a hearing aid that could pick out one speaker from a group.
One-on-one conversation with Voice AI? Easy. Multi-speaker chaos? That’s the real challenge. 🎙️
We used Fishjam’s Selective Forwarding Unit to help Gemini Live API handle interruptions and cross-talk in our mystery game, Deep Sea Stories.
Learn more – full blogpost in 🧵