NVIDIA Releases NemotronLabs VoiceChat 11B: An Open Full-Duplex Speech-to-Speech Model with ~450 ms Turn-Taking and Live Tool Calling
NVIDIA reports this unified 11B model achieves 448 ms latency by removing cascaded API handoffs.
Daily edition
NVIDIA releases NemotronLabs VoiceChat 11B for full-duplex speech.
Daily edition
50 of 50 feeds responded
198 distinct stories
38 previously issued, held back
7 worth your time
Collection and selection
198 candidate events; 38 shortlisted stories; 31 with article text available; 7 selected. Counts describe different stages; clustering, validation and refills can change the shortlist.
Active filters: none. Sort: Ranked.
All 7 stories, in ranked orderMore filters
GitHub Models is now retired
GitHub has completed the retirement of its unified model playground and API.
Quoting Claude Opus 5 system prompt
Anthropic suspended and later restored access to Claude Fable 5 and Mythos 5 due to export controls.
AI assistant hacks gym website in first known Australian autonomous cyber attack
An AI assistant reportedly exploited a booking vulnerability to reserve a gym class months early.
Also covered bySimon WillisonNEC tests parking tech that only starts charging once you exit your car
NEC is testing parking technology that initiates billing only after the driver exits the vehicle.
An OpenAI Strategist Says AI Labs Should Rival Government Power
An OpenAI strategist argues AI labs should eventually rival government power.
Making an AI bid writer refuse to lie
A user configured an AI bid writer to identify and refuse to lie in tender documents.
No stories match. Active filters.