Researchers Artem Saloev, Erin Pacquetet, and Nicolas Ballier found that the ABX experiment with Mimi's semantic token codebook fails to capture the mapping of semantic tokens to phone realisations, b…
OpenLive v26.7.15, an open-source runtime for continuous voice agents, delivers a GPT-Live-comparable experience with WebRTC, semantic endpointing, and a model-neutral architecture. The project, which…
Real-time AI interaction models like Moshi and Qwen-Omni face a hidden collapse risk due to swelling KV cache, causing sudden performance drops under sustained load. A simple fix—bounding each session…
Rudrite Research released a 47-minute documentary explaining how modern text-to-speech works, from sound waves as numbers to neural audio codecs and talking LLMs. The film covers key milestones like W…
Developer Jeff Triplett describes a remote development setup using Tailscale, tmux, and Mosh to seamlessly move between devices without losing work. The stack enables secure, persistent sessions acros…
Amazon launched Sleep Studio on Wednesday, a new feature for Echo smart speakers designed to help parents manage children's bedtime routines. The feature includes bedtime stories, relaxing sounds, and…
Thinking Machines released TML-Interaction-Small, a 276B parameter mixture-of-experts model with 12B active parameters, advancing real-time voice interaction by processing audio and images in under 20…