Daily Brief Archive
Explore our complete collection of AI insights. Every edition, every breakthrough, all in one place.
DeepSeek reportedly plans to launch V4 around the Lunar New Year, focusing on coding and ultra-long code prompt handling; Lightricks open-sources LTX-2 (19B parameters) supporting synchronous 4K video and audio generation, up to ~20 seconds; TrendForce says HBM4 mass production may be delayed until after late Q1 2026 due to NVIDIA raising Rubin memory specifications.
Google pushes Gemini summaries/suggested replies to over 3 billion Gmail users and enables them by default for some, requiring manual opt-out; AI21 open-sources the Jamba2 family under Apache 2.0, offering 256K context and 3B/52B variants; Hyundai plans annual production of 30,000 Boston Dynamics Atlas humanoid robots at a US factory by 2028
Mobileye plans a $900 million acquisition of Mentee Robotics expected to complete in Q1 2026; Zhipu AI discloses Hong Kong IPO with an issue price of HK$116.2 per share, raising about HK$4.3 billion and valuing the company over HK$50 billion; malicious Chrome extensions with over 900,000 combined downloads steal and exfiltrate ChatGPT/DeepSeek conversation content.
xAI announced the completion of a $20 billion funding round with NVIDIA's participation to expand Grok and build its own AI infrastructure; Accenture acquired UK AI company Faculty for $1 billion and appointed its CEO as Group CTO; Open WebUI patched the high-risk CVE-2025-64496 vulnerability (affecting up to v0.6.34) to block SSE injection and potential RCE risks
NVIDIA open-sources Alpamayo 1 (10-billion parameters) and provides 1700+ hours of driving data and the AlpaSim simulation framework; NVIDIA releases Jetson T4000 (1,200 FP4 TFLOPS, 64GB memory) priced at $1,999 per unit for orders of 1,000 or more; TII open-sources Falcon H1R 7B and discloses AIME 2025 score of 83.1
Global GPU order volume expected to be ~2 million but immediate supply is only ~700,000, prompting large companies to turn to ASICs; LMArena completes $100M funding round to support user-vote-driven model evaluation and inference costs
China Meteorological Administration released/upgraded 5 major weather AI models; Fengqing generates 15-day global forecasts in 3 minutes, Fenglei's strong echo forecast quality improved over 25%; xAI's Grok is accused of generating sexualized 'undressing' images, Reuters counted at least 21 requests fully satisfied, and India demands a rectification report within 72 hours; Biren Technology raised $558 million in its HK IPO, surged 76% on the first day to reach a market cap of HK$46.9 billion.
Google releases Gemini 3 Pro, achieving 1501 Elo on LMArena, and Deep Think improves Humanity’s Last Exam score to 41%; ByteDance plans to procure approximately $14 billion worth of Nvidia chips in 2026, with Chinese companies booking over 2 million H200s; Nvidia reportedly acquires inference chip company Groq for $20 billion.
OpenAI plans to release a new-generation audio model in Q1 2026, paving the way for screenless personal devices; Z.ai open-sources GLM-4.7, achieving 67.5 on BrowseComp and 87.4 on τ²-Bench, and launches API; DeepSeek proposes mHC architecture, validated on 3B/9B/27B models with training overhead increasing only 6.7% at scaling factor 4.