#Ai

Nvidia will pay $6 billion for a license to Poolside's technologies and poach 109 of the startup's employees, investing an additional $1 billion in its business. The goal is to create an open AI model to compete with Chinese counterparts.

OpenAI is preparing a Locked Chats feature for the Android version of ChatGPT: conversations will be hidden from the general history and opened only via a PIN code or fingerprint.

Pew Research: 35% of web pages published after ChatGPT's launch were written with AI involvement. In the overall sample the share is about 10%, while commercial .com domains lead in AI content.

Japan is actively rolling out facial recognition: from payments at Trial stores and turnstiles on the Tōjō Line to access at Tokyo Dome. The biometrics market could grow to 130 billion yen by 2030, but privacy concerns are rising as well.

Scientists at the University of California, San Diego used AI to decode the start sequence of human DNA, analyzing around 500,000 promoter variants and paving the way for early cancer prediction.

AVIC Chengdu engineers, developers of the J-20 and J-36 stealth fighters, warned in a scientific paper that AI may embed critical defects in military aviation due to inaccurate data.

Google Research showed that LLMs encode 95–98% of facts but fail to retrieve 26–34% of them. The new knowledge profiling method separates real knowledge gaps from access failures, including the word-order paradox.

Xiaomi reported its second quarter of 2026: revenue fell 7% to $16.2 billion, smartphone shipments dropped 26.3%, while the EV and AI segment grew 17%.

Google has set September 15, 2026, for the Googlebook laptop presentation: the search giant will reveal the specifications, partners, and operating system of the next-generation devices.

🤖 Anthropic CEO Dario Amodei explains why society doesn't trust AI: it's a trust crisis, not a real threat. He believes AI must prove its value through real achievements, such as curing cancer. Industry regulation must be balanced to ensure safety and innovation. #AI #Anthropic #DarioAmodei #TrustCrisis #CancerTreatment

🇨🇳 Chinese Z.ai unveils GLM-5.3 — a powerful AI model with open weights capable of competing with Anthropic and OpenAI. 🚀 #AI #Zai #GLM53 #OpenAI #Anthropic

🤖 The 2026 World Humanoid Robot Games have kicked off in Beijing! 🇨🇳 China dominates: 1,975 robots from 641 teams vs. 85 teams from 15 countries. 📈 51 competitions, including running, tennis, and tool assembly. 🚫 Ban on remote control: robots must act autonomously. 🏭 Goal: Turn medals into orders and deploy robots in real workplaces. #robots #technology #china #2026

🚀 Anthropic is preparing for an IPO with a $2 trillion valuation — this could become the largest offering in history, surpassing SpaceX. But can the company justify such expectations? 💰 Revenue is growing, but profit is still far from the required figures. Investors believe in the future, but critics warn of a bubble. What happens next? 🤔 #AI #IPO #Anthropic #Claude #TechNews

🇨🇳 China is changing its strategy to combat climate disasters. After summer floods that claimed dozens of lives in Gansu and Chongqing, Xi Jinping ordered the implementation of AI and drones for disaster monitoring. Beijing has been identified as a risk zone due to outdated drainage infrastructure. #China #Climate #Floods #XiJinping #Technology

🤖 In China, a Unitree G1 humanoid robot is asking passersby for money for 'charging'. The video of the robot kneeling with a QR code has shocked the internet. Is this marketing, street art, or the beginning of a new era? #Unitree #Robots #China #Technology

🚀 Google wins in court: Rospatent allows the registration of the Gemini brand in Russia! 🇷🇺 After 2 years of dispute with Czech company Finshape, the agency recognized visual differences and allowed registration in the AI services class. This is a step towards the legal launch of AI bots in the RF! 🤖 #Google #Gemini #Rospatent #AI #Technology

🚨 Zoomsday: Critical Zoom Vulnerability Created by AI in 20 Prompts Allowed Hackers to Hijack Devices Without Victims' Knowledge. A Security researchers showed how AI can automate exploit creation. Zoom has already released a fix, but the incident confirms experts' fears: cyberattacks are becoming more accessible. #Zoom #CyberSecurity #AI #Zoomsday

🇦🇲 The largest AI factory in the CIS has opened in Armenia! Firebird built a data center in Hrazdan in six months with NVIDIA support. 🚀 6,144 B200 GPUs are already operational, with plans for 70,000 chips by 2027. 💰 Investments reach $4 billion. #AI #Armenia #NVIDIA