OpenAI launches GPT-Live, a full-duplex voice AI that listens and speaks simultaneously for more natural conversations, but its strikingly human-like tone raises serious concerns that users may too easily trust its hallucinations.
Luma COO Caroline Ingeborn declares that true AGI demands omnimodal AI systems capable of processing text, video, images, and audio simultaneously, as the company sets its sights on revolutionizing creative industries while insisting the technology will empower — not replace — human creativity.
NVIDIA-backed Luma AI reaches a $4 billion valuation after securing $900 million in Series C funding led by Saudi-owned Humain, and is now expanding to London with 200 new hires as its Ray3 world model reportedly outperforms OpenAI's Sora and rivals Google's Veo 3.
WisprFlow is revolutionizing Mac productivity by enabling users to dictate text at 179 WPM using AI-powered transcription that removes filler words, corrects grammar, and adapts tone across all apps — though its macOS exclusivity, 6-minute cap, subscription cost, and cloud-based processing present real trade-offs.
AI SDK 7 launches with groundbreaking features including reasoning control, durable agents, real-time voice support, and experimental video generation, solidifying its position as a powerhouse tool for AI development with over 16 million weekly downloads.
NVIDIA launches an open source XR AI beta platform enabling developers to build real-time visual and voice AI agents for AR glasses and XR headsets, powered by GPU-accelerated services including Cosmos, Nemotron, and NeMo Agent Toolkit.
Google launches Gemini 3.5 Live Translate, a groundbreaking AI model delivering near real-time speech-to-speech translation across 70+ languages while preserving speakers' natural voice characteristics, now available to developers, enterprise users, and the public via Google Translate on Android and iOS.
Google's Magenta team releases Magenta RealTime 2, a cutting-edge open-weights AI music generation model available in 230M and 2.4B parameter sizes, enabling real-time music streaming on Apple Silicon Macs and offline inference on NVIDIA GPUs.
Miso Labs launches MisoTTS, an open-source 8-billion-parameter text-to-speech model now live on Hugging Face, featuring voice cloning, highly emotive conversational speech, and built-in audio watermarking for safety.
Google DeepMind launches Gemma 4 12B, a multimodal AI model that runs locally on consumer laptops with just 16GB of VRAM, featuring a groundbreaking encoder-free architecture that processes vision and audio directly through its LLM backbone for lower latency and memory use, available now on Hugging Face and Kaggle under …