May 19, 2026United StatesSaveSavedSaveGoogle Releases Gemini 3.5 Flash Model: Output Speed 4x Faster Than GPT-5.5
May 19, 2026United StatesSaveSavedSaveGoogle Gemini Integrates C2PA Content Credentials to Help Identify AI-Edited Content
May 19, 2026United StatesSaveSavedSaveGoogle Releases Gemini Omni All-Round Model: Supports Multimodal Processing & Conversational Editing
May 19, 2026United StatesSaveSavedSaveGoogle Processes Over 3.2 Quadrillion Tokens Monthly, Up 7x YoY
May 19, 2026Mainland ChinaSaveSavedSaveAMD AI DevDay Makes Its First Appearance in China, Showcasing Edge Agent and Open Source Platform Progress
May 19, 2026Mainland ChinaSaveSavedSaveLenovo Launches Tianxi AI 4.0: Adds Local Bionic Memory, Personal Knowledge Base and Skills Plaza Features
May 19, 2026Mainland ChinaSaveSavedSaveDeepSeek Responds to Abnormal Special Character Triggers: Model Hallucination, No Privacy Leak
May 19, 2026Mainland ChinaSaveSavedSaveZhang Chaoyang on AI in Science Popularization: Convenient but Needs Guidance; Over-reliance Shrinks Thinking
May 19, 2026Mainland ChinaSaveSavedSaveUnitree Robotics Releases G1 Robot Voice Control Video: AI Generates Real-Time Actions in One-Take Footage
May 19, 2026United StatesSaveSavedSaveFigure AI Robot vs Human in Package Sorting: Human Narrowly Wins, CEO Says It's the Last Time
May 19, 2026United StatesSaveSavedSaveAMD's Lisa Su: 5 Billion People to Use AI Daily in 5 Years; Right Apps Are Key
May 19, 2026Mainland ChinaSaveSavedSaveHorizon Open-Sources 400M-Parameter Robot Cerebellum Model HoloMotion-1, Enabling Various Complex Actions
May 19, 2026Mainland ChinaSaveSavedSaveTencent Cloud's Hy3 preview and DeepSeek-V4-Pro models to end free public beta on May27, switch to commercial service
May 19, 2026Mainland ChinaSaveSavedSaveState-owned institutions and central enterprises join Kimi's shareholder list; financing wraps up with valuation quadrupling
May 19, 2026BelgiumSaveSavedSaveBelgium's Imec Unveils 3D CCD Memory Architecture, Combining DRAM Speed and NAND Density to Boost AI Inference Performance