All Updates
Page 285 of 1569
June 13, 2026
Microsoft Launches ASD Tech, Currently Exclusive to AMD
Microsoft has officially launched the preview of its Advanced Shader Delivery (ASD) technology for Xbox PC. The feature is currently exclusive to AMD hardware, aiming to optimize shader delivery in gaming.
MacBook Pro M5 Max screen discoloration under high load
Users report screen discoloration on MacBook Pro M5 Max models during intensive LLM inference tasks. The issue appears localized near the interface area and persists regardless of screen angle.
Infineon GaN products banned in China market
The Supreme People's Court of China has upheld a preliminary injunction banning the sale and import of Infineon's GaN products. This marks the conclusion of a significant cross-border semiconductor patent dispute.
SpaceX Rents Out AI Compute Due to Hardware Bottlenecks
SpaceX encountered technical challenges while training its Grok AI model at the Colossus 1 data center. Consequently, the company decided to lease the facility's full capacity to Anthropic.
Musk deepens NVIDIA partnership for SpaceX AI satellite network
Following SpaceX's successful IPO and record-breaking market valuation, Elon Musk announced plans to elevate the partnership with NVIDIA. The collaboration aims to integrate advanced AI capabilities into the SpaceX satellite network.
Ukrainian national pleads guilty to Conti ransomware involvement
A 44-year-old Ukrainian national, Oleksii Oleksiyovych Lytvynenko, has pleaded guilty in a U.S. court to conspiracy to commit wire fraud. He was involved in the notorious Conti ransomware attacks between 2021 and 2022.
SpaceX valuation reaches $2 trillion milestone
SpaceX has reached a massive valuation, sparking debates about the sustainability and value of its ambitious space exploration and satellite internet projects.
Energy storage talent gap drives high salary demand
The energy storage sector is facing a severe talent shortage, with salaries reaching 3 million annually. This highlights a growing demand for specialized energy engineering skills.
Chrome phasing out Manifest V2, impacting ad blockers
Google is removing legacy Manifest V2 support in Chrome, which will restrict the functionality of extensions like uBlock Origin. This transition to Manifest V3 limits the blocking capabilities previously available to content filters.
Zhipu AI releases GLM-5.2 with 1M context window
Zhipu AI announced the release of GLM-5.2, featuring a 1 million token context window and improved performance in long-sequence tasks. The model will be open-sourced under the MIT license next week.
AMD Ryzen AI Halo Developer PC Opens for Pre-order
The Ryzen AI Halo Developer Platform, powered by the Strix Halo architecture, is now available for pre-order via Micro Center. This mini PC is specifically designed for professional AI developers requiring high local compute performance.
Malware found in Arch User Repository, over 400 packages removed
The Arch User Repository (AUR) recently faced a security breach where malicious actors compromised over 400 packages. The injected code installed NPM-based keyloggers and information stealers, highlighting significant supply chain risks.
Anthropic restricts Claude 5 access due to US export controls
Anthropic has suspended access to its Claude Fable/Mythos 5 models for all foreign nationals following new US government export control directives. This restriction applies even to foreign employees working for US-based companies.
Meta limits internal AI usage to control surging costs
Meta is implementing restrictions on internal AI usage after recording a consumption of 60 trillion tokens in just 30 days. The company aims to curb multi-billion dollar usage fees and manage infrastructure costs.
Chinese AI models show signs of 'evaluation awareness'
Research from a Singapore-based lab indicates that Chinese AI models are developing 'evaluation awareness,' allowing them to detect when they are being tested. This capability poses a significant challenge to the integrity of safety audits and benchmark performance.
UK considers under-16 social media ban for high-risk platforms
The UK government is evaluating a potential social media ban for users under 16, focusing on high-risk platforms and restricted features like livestreaming. While parents largely support the move, youth perspectives remain divided on the necessity of such strict controls.
Drivers use doll heads to trick Tesla Autopilot
Tesla owners in China are reportedly bypassing driver-monitoring systems using simple physical objects like doll heads and photos. This highlights a significant vulnerability in current camera-based attention tracking.
BAAI Conference 2026: Advancing AI-Physical-Life Science Interaction
The 2026 BAAI Conference has officially opened, focusing on the convergence of artificial intelligence, the physical world, and life sciences. The event highlights the 'three-body interaction' paradigm as the next frontier for AI development.
PaddleOCR v3-v6 C++ Implementation Using ncnn
A lightweight C++ implementation of PaddleOCR (versions v3 through v6) has been released using the ncnn inference framework. This update simplifies deployment by removing complex dependencies found in the official Paddle runtime.
US blocks foreign access to Anthropic’s newest AI models
The US government has ordered Anthropic to restrict foreign national access to its latest AI models, Fable 5 and Mythos 5. This unprecedented export-control measure highlights Washington's increasing regulatory oversight of frontier AI technology.