Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

24 Aug 2026

AI Drone With Nvidia Chip Kills Three Civilians in Ukraine's First Fully Autonomous Strike

A Russian Molniya drone powered by an Nvidia Jetson Orin module killed three civilians at a Zaporizhzhia gas station in July after selecting its final target autonomously. Ukrainian forensic teams confirmed it's the first documented case of civilian deaths caused by a Russian drone using fully autonomous targeting, raising alarm about AI-powered systems making life-or-death decisions without human intervention.

6 Sources

Thomson Reuters Unveils $40M Proprietary AI Model Built on Alibaba's Qwen to Power Legal Work

Thomson Reuters launched its first proprietary large language model called Thomson, built on Alibaba's Qwen foundation for $40 million. The in-house AI model powers CoCounsel Legal to reduce reliance on Anthropic while giving the company full intellectual property control over its AI assistant for lawyers.

5 Sources

IBM Unveils Dual-Architecture Chip Running Arm and Z Instructions at 5.7 GHz on 2nm Process

IBM revealed its first dual-architecture chip at Hot Chips 2026, featuring 11 cores built on 2nm process running at 5.7 GHz base frequency. Each core natively executes both Arm and Z instructions, switching between architectures in nanoseconds. The processor brings Arm's 22 million-developer ecosystem to mainframes while maintaining mission-critical reliability for enterprises.

9 Sources

Alibaba Launches Wan3.0 AI Video Model Days After $10.2 Billion Share Sale to Fund AI Infrastructure

Alibaba unveiled Wan3.0, its latest AI video generation model capable of creating 30-second clips from diverse inputs including PDFs and PowerPoint files. The launch follows a record $10.2 billion Hong Kong share placement, with all proceeds earmarked for AI capabilities. Priced competitively against Google Veo, the model targets corporate and marketing use cases while raising questions about profitability amid soaring AI infrastructure costs.

5 Sources

STMicroelectronics and National University of Singapore Launch Edge AI Lab to Advance Embodied Intelligence

STMicroelectronics and the National University of Singapore unveiled the ST-NUS HELIX Corporate Lab, a four-year research initiative focused on edge AI technologies. The lab will develop system-to-silicon innovation for generative and embodied AI applications using 18nm FD-SOI technology and embedded Phase Change Memory to address energy efficiency challenges in AI hardware.

2 Sources

Sonova Unveils Phonak EON: Third-Gen AI-Powered Hearing Aid With 2x Speech Clarity

Sonova launched Phonak EON, its third real-time AI hearing-aid platform featuring the HYPERSONIC processor that cuts power draw by 37%. The device makes speech twice as understandable while being 25% smaller and 20% lighter than previous models. This launch aims to help Sonova meet medium-term growth targets that faced investor skepticism in March.

4 Sources

Claude Down: Anthropic's AI Platform Hit by Multi-Model Outage Affecting Thousands

Anthropic's Claude AI service experienced a major outage on August 24, affecting multiple models including Claude Opus 5, Claude Mythos 5, and Claude Fable 5. Over 1,300 users in the US and 180 in India reported issues with Claude down, primarily impacting Claude Code. The company identified elevated errors and deployed engineers to restore service.

5 Sources

Windows 11 Tests User Control Over Unified Memory Split for AI Workloads and Gaming

Microsoft is testing IntelligentCarveout in Windows 11 build 29648.1000, a hidden feature that lets users manually reserve unified memory for graphics and AI tasks. The experimental functionality could give RTX Spark and AMD Ryzen AI Max systems more predictable performance by protecting GPU memory from competing applications.

5 Sources

23 Aug 2026

ChatGPT for iPhone Adds Photo Feature as OpenAI Targets EU's 450 Million Users Without Siri AI

OpenAI introduced a long press feature in ChatGPT for iPhone that surfaces the four most recent photos instantly. Released on August 21, the update streamlines adding images to prompts while positioning OpenAI to fill the visual assistant gap left by Apple's delayed Siri AI rollout in the European Union.

4 Sources

Z.ai Confirms Ox Alpha is GLM-5.3-Flash After Anonymous Launch Tops OpenRouter Charts

Chinese AI company Z.ai confirmed it developed Ox Alpha, the mysterious free AI model that dominated OpenRouter last week. Now officially called GLM-5.3-Flash, the open-weight model excels at coding and agentic work but raises data privacy questions as prompts are retained by the provider.

16 Sources

22 Aug 2026

ShieldFont Disrupts AI Scrapers by Poisoning Unauthorized Data Collection with Hidden Word Swaps

Designers created ShieldFont, a font that looks normal to humans but disrupts AI scrapers by swapping words in HTML source code. The innovation poisons scraped data, making unauthorized LLM data collection costlier while raising questions about accessibility.

2 Sources

Humanoid Robots Shatter Usain Bolt's 100-Meter Record, Clocking 8.64 Seconds at Beijing Games

China's Tiangong Ultra humanoid robot ran 100 meters in 8.64 seconds at the World Humanoid Robot Games in Beijing, beating Usain Bolt's 9.58-second record. The event showcased rapid advancements in robotics but also revealed significant challenges in high-dexterity tasks like folding laundry and using tools.

17 Sources

DeepSeek Launches Multimodal AI Model Approaching Anthropic Opus 4.8 Performance at Fraction of Cost

DeepSeek unveiled DeepSeek-V4-Flash-Vision-Exp, an experimental multimodal AI model with image understanding capabilities that approaches Anthropic's Opus 4.8 performance. The model wins on three benchmarks including Agents' Last Exam and ZeroBench while maintaining 99% lower operational costs, intensifying competition in the global AI landscape.

3 Sources

Nvidia proves AI harness, not model, drives agent performance with 100% ARC-AGI-3 score

Nvidia's research reveals that its AVO coding agent achieved a perfect 100% score on the ARC-AGI-3 benchmark using Claude Opus 5, while the model alone scored just 30%. The breakthrough demonstrates that the AI harness—the wrapper around the model—matters far more than the underlying AI for long-horizon tasks.

2 Sources

Linus Torvalds Uses AI to Fix Linux Bug After Grueling Debug Session From Hell

Linux creator Linus Torvalds wrestled with an AI coding assistant through what he called a debug session from hell to fix a Linux kernel bug. The AI repeatedly declared the problem impossible and unsolvable, but Torvalds persisted through 24 patches and 18 kernel boots to narrow down the issue to a single line of code.

3 Sources

Get all updates into your inbox 💌, Stay Up-to-date for Free 🥳