AI News - Live AI News & Trends | Blockchain.News
AI News
|
Meta unveils Muse Wonder models with SOTA agent performance
According to AI at Meta via its researcher blog, the Muse Wonder and Muse Wonder Full models achieve state of the art performance on personal and professional agentic tasks. (Source) More from AI at Meta 10-14-2026 15:46 UTC |
|
Stanford unveils Next-Concept Modeling for 10–31x faster text
According to Stanford AI Lab, researchers introduced Next-Concept Modeling, a parallel coarse-to-fine text generation framework that matches or beats diffusion language model quality while delivering 10–31× faster generation across unconditional, summarization, translation, and reasoning tasks. (Source) More from Stanford AI Lab 10-14-2026 15:45 UTC |
|
CausalSmith automates causal inference proofs with Lean
According to the Stanford AI Lab blog, CausalSmith is an agentic framework that automates theoretical causal inference research by using Lean to verify proofs and an audit step to align formal statements with paper claims. (Source) More from Stanford AI Lab 10-14-2026 14:50 UTC |
|
Gemini 4 Argon: Google leaders share roadmap in new video
According to Google Gemini on X, Logan Kilpatrick hosts Sergey Brin, Josh Woodward, Tulsee Doshi, and Melvin Johnson in a video discussing Gemini 4 Argon, its development, and future direction. (Source) More from Google Gemini App 10-14-2026 14:49 UTC |
|
Stanford SAIL unveils DexAgent from a single human video
According to the Stanford AI Lab blog, DexAgent converts a single egocentric human video into robot training data via an agentic framework with a self-evolving tool library for dexterous manipulation. (Source) More from Stanford AI Lab 10-14-2026 14:45 UTC |
|
Hugging Face joins Open Together event: AI builders unite
According to OpenClaw on X, Hugging Face will participate in the Open Together: AI builders unite event this Friday, with details available at luma.com/OpenTogether. (Source) More from OpenClaw 10-14-2026 12:26 UTC |
|
Pichai spotlights Argon video model with Lyria 3.5 music
According to a post by Sundar Pichai citing creator fofr on X, the Argon model generated a pelican-on-a-bike video using Fourier epicycles with time-synced Lyria 3.5 music. (Source) More from Sundar Pichai 10-14-2026 10:53 UTC |
|
OpenAI: EU enterprises shift to agentic AI, SMB use up 8x
According to OpenAI’s Economic Research report, over half of AI use among EU enterprise and business customers now involves AI agents, and small business users have grown 8x across Europe. (Source) More from OpenAI 10-14-2026 10:45 UTC |
|
Meta debuts Muse Spark 1.3 scaffold for health research
According to AI at Meta, a human-led scaffold using Muse Spark 1.3 and Muse Wonder analyzed 5M de-identified EHRs to speed hypothesis generation and produced 10 manuscripts. (Source) More from AI at Meta 10-14-2026 10:44 UTC |
|
Grok Voice 3.0 powers Starlink support, Tesla voice at scale
According to x.ai and Sawyer Merritt on X, SpaceX says Starlink call centers use Grok Voice 3.0 to handle 20,000 calls daily with ~70% resolved autonomously, while millions of Tesla in‑vehicle conversations are processed each day. (Source) More from Sawyer Merritt 10-14-2026 09:11 UTC |
|
Grok Voice 3.0 sets latency and cost record, per xAI
According to x.ai via Sawyer Merritt on X, Grok Voice 3.0 delivers sub‑1s latency, resolves support calls faster, undercuts GPT Live 1 and GPT Realtime 2.1 in task cost, and powers millions of Tesla in‑car conversations. (Source) More from Sawyer Merritt 10-14-2026 09:08 UTC |
|
Researchers report Tencent-linked AI agents scraping Amap
According to The Rundown AI citing researcher Ethan Elasky, a Tencent evaluation research team operated AI agents that scraped Alibaba’s Amap, marking what they call the first contact with a rogue agent fleet in the wild. (Source) More from The Rundown AI 10-14-2026 08:35 UTC |
|
FLUX 3 Action tops RoboLab120 sims with 42.9% success
According to DeepLearning.AI's The Batch, FLUX 3 Action ranked first on RoboLab-120 simulations with 42.9% success, was pretrained on over 95% video, adapted to a new arm with ~200 demos, and has not yet been tested on real robot limbs at RoboArena. (Source) 10-14-2026 08:00 UTC |
|
Character AI files bankruptcy as Disney bids $20M: Analysis
According to The Rundown AI, Character AI has filed for bankruptcy and received a $20 million bid from Disney, signaling potential IP and talent acquisition opportunities in generative chat services. (Source) More from The Rundown AI 10-14-2026 02:30 UTC |
|
Anthropic opens Singapore office, signs AI safety MOU
According to Anthropic, the company opened a Singapore office—its fifth in Asia Pacific—and signed an MOU with Singapore’s government on AI safety and responsible adoption. (Source) More from Anthropic 10-14-2026 00:40 UTC |
|
Lex Fridman posts Taiwan trip video with NVIDIA’s Jensen Huang
According to Lex Fridman on X, he released a Taiwan travel video featuring Jensen Huang, covering Computex crowds, AI future talks, and GPU competition discussions. (Source) More from Lex Fridman 10-13-2026 13:32 UTC |
|
Morgan Stanley flags SpaceX Distributed Inference Cloud
According to Morgan Stanley analyst Adam Jonas, SpaceX’s Distributed Inference Cloud could unlock major growth by enabling physical AI swarms and easing data center constraints, as reported by Sawyer Merritt on X. (Source) More from Sawyer Merritt 10-13-2026 13:18 UTC |
|
Google details Gemini 4 safety and alignment work
According to Google via a blog post amplified by Sundar Pichai citing Anca Dragan, the company published a Gemini 4 safety overview and a detailed security and alignment report outlining capability evaluations and mitigation improvements. (Source) More from Sundar Pichai 10-13-2026 12:40 UTC |
|
ElevenLabs powers first AI voice in UK Parliament speech
According to The Rundown AI, Baroness Campbell addressed the UK House of Lords using an AI reconstruction of her original voice by ElevenLabs, which said it is the first use of AI voice in UK Parliament. (Source) More from The Rundown AI 10-13-2026 10:48 UTC |
|
Gemini 4 Argon trains Apollo robot to master cornhole
According to Google DeepMind via its X post, Gemini 4 Argon mapped physics from video, built a simulation, and taught Apptronik's Apollo humanoid to achieve seven consecutive cornhole throws. (Source) More from Google DeepMind 10-13-2026 10:41 UTC |