CATEGORY_INDEX // 237 ENTRIES

Artificial Intelligence

Artificial intelligence, machine learning, LLMs, generative AI, and AI platform news.

FILTER: Consumer Tech & Hardware Gaming Xiaomi Mobile & Smartphones Chips & AI AI SORT: NEWEST ▾

Diagram titled Adreno Neural Fusion showing stacked chip layers labeled Qualcomm Oryon CPU, Adreno GPU, and Hexagon NPU, illustrating their integration

CHIPS & AI · 09/02/2026

Adreno Neural Fusion Puts Matrix Cores in the GPU

Adreno Neural Fusion adds AI matrix units to Qualcomm's GPU a year after Apple and level with Arm, but the 18MB of HPM is…

7 MIN

READ →

Illustration titled "Nvidia Buys Hugging Face, $12.9B After the Breach," depicting an unconfirmed acquisition with model migration status, security breach warnings, and open-weight model listings like Meta-Llama and Mistral-Nemo, from m1k.tech

AI · 09/02/2026

Nvidia Just Agreed to Buy the Company Its Own AI Agent Broke Into

An OpenAI agent infiltrated Hugging Face's system, minting credentials and enrolling 181 devices undetected for over four days. Nvidia plans to acquire Hugging Face…

4 MIN

READ →

Diagram of Sonos 27 MCP AI-native audio platform: Ace Ultra headphones and Beam Ultra soundbar connect to an MCP AI Server, routing ChatGPT (GPT-4o) and Gemini (2.5 Pro) through custom agents (Soundscape Curator, Listening Advisor) for room-aware, adaptive audio experiences

AI · 09/01/2026

Sonos Is Selling You an MCP Server for Your Living Room

Sonos unveiled its "Sonos 27mcp" protocol, enabling integration of ChatGPT and Gemini into its audio systems, alongside the new Ace Ultra headphones and Beam…

4 MIN

READ →

Infographic: Alibaba's $10.2B Hong Kong share placement equals about 1/4 of its ~$40.8B quarterly AI capex burn; Qwen inference costs estimated $1.9B+/day

AI · 08/29/2026

Alibaba Raised $10.2 Billion in an Afternoon

Alibaba's share placement raised $10.2 billion, the largest primary follow-on in Hong Kong history. Against RMB67.7 billion of June-quarter capex, it buys about one…

8 MIN

READ →

Infographic on Anthropic's Custom Silicon initiative: TPU heritage across 7 generations, efficiency trajectory chart, next-generation accelerator concept with scalable interconnect, domain-specific cores, optimized memory hierarchy, and system-level co-design; institutional groundwork covering long-term infrastructure strategy, vertical integration, model-silicon co-design, and supply chain resilience; headline "The TPU Architect Hire" noting the architect of seven TPU generations now building what's next

AI · 08/25/2026

Anthropic Hired the Person Who Built Google’s TPU Program

Seven. That’s how many generations of Google’s Tensor Processing Unit Amir Salek shipped before he left in 2022, and it matters more than his…

4 MIN

READ →

Infographic: EU industrial rules cover physical goods and production but not software or digital design, leaving a digital sovereignty gap with code, CAD files, and cloud infrastructure falling outside EU jurisdiction

DEEP TECH · 08/24/2026

Europe’s New “Made in EU” Law Never Mentions Software

The EU's Industrial Accelerator Act mandates that by January 2029, various products, including buildings and vehicles, must contain a certain percentage of Union-origin materials.…

5 MIN

READ →

Infographic: Alibaba's Qwen3.8 Max, a 27B-parameter open-weight AI model, runs on-device via NPU at ~148 tokens/sec, Apache 2.0 licensed, contrasted with Meta's closed-weight Llama

AI · 08/23/2026

Qwen3.8 Max Is Now Open Weights, and Alibaba Beat Meta to the Laptop Model

Alibaba has launched Qwen3.8 Max's weights and released Qwen3.8-27B, a laptop-optimized model claiming to match larger models' performance. This development responds to Meta's recent…

3 MIN

READ →

Diagram of Mistral Large 3: European AI platform using DeepSeek-inspired MHLA/MoE architecture, hosted on GLM-5.2 for sovereign, transparent AI infrastructure

AI · 08/22/2026

Mistral Large 3 Runs DeepSeek’s Blueprint. Now Mistral Hosts Chinese Weights Outright.

Mistral Large 3 shares DeepSeek V3's attention stack down to the compression ranks, and since 11 August Mistral serves Z.ai's GLM-5.2 unmodified on its…

10 MIN

READ →

AI · 08/21/2026

Cerebras Launched a Faster Chip the Same Week It Admitted the Business Still Loses Money

Cerebras unveiled the CS-4, a powerful server rack featuring three wafer-scale chips, promising significant throughput improvements. However, despite reported advancements and a $180.1 million…

4 MIN

READ →

Infographic on Ryanair x Google Cloud: Gemini Enterprise AI powers flight scheduling, customer service (1.2M+ daily conversations, 92% resolution, under 60s response), and real-time operations (99.7% on-time, 18% fuel savings, 500+ daily flights), contrasted with a sovereign European cloud route offering limited scale, higher cost, and slower innovation. Tagline: All In on Gemini Enterprise – Scale, Speed, Intelligence, Winning Airline Technology

AEROSPACE · 08/20/2026

Ryanair Just Went All In on Google Cloud, Right as Airbus Walked the Other Way

Ryanair has struck a five-year deal with Google Cloud and its suite of tools to support its workforce of 35,000 employees, aiming to double…

3 MIN

READ →

Infographic comparing Microsoft's AI claims vs reality: 5.0+ GW power capacity claimed across 50+ campuses, but only ~180K-220K GPUs (10-15% of claimed potential) actually operational, mostly H100 and A100 clusters, showing a gap between empty warehouse shells and running data centers

AI · 08/20/2026

Microsoft’s AI Chip Count Doesn’t Match Its Own Numbers

A Guardian investigation into internal Microsoft documents found the company has 2.2 million AI chips installed, a fraction of what its own gigawatt claims…

5 MIN

READ →

Infographic on Sony PSSR 2.0: compares traditional GPU upscaling pipeline to new kernel-predicting neural network, showing 540p input upscaled to 2160p output with 90% less compute time, edge detection and noise separation, GPU reconstruction, and PS5 real-time graphics optimization

AI · 08/19/2026

Sony Cut PSSR 2.0’s Training Time From Four GPU-Months to One GPU-Week

Sony's PSSR 2.0 significantly reduced training time from four GPU-months to under one week by simplifying its neural network's functions. It switched from a…

4 MIN

READ →

AI · 08/18/2026

Washington’s Pax Silica Letter Turns AI Diplomacy Into a Loyalty Test

The U.S. State Department's recent letter urges 35 countries to choose sides in the AI competition between Washington and Beijing, highlighting concerns over Kazakhstan's…

5 MIN

READ →

Chicago Mayor Brandon Johnson's executive order defining data centers leaves the definition of "data center" intentionally blank, with no definition provided

CHICAGO · 08/17/2026

Johnson’s Data Center Order Doesn’t Define a Data Center. He Used It as a Guarantee Anyway.

Chicago's new executive order on data centers lacks a clear definition of what constitutes a data center, complicating regulatory measures. Mayor Brandon Johnson faces…

9 MIN

READ →

Infographic: AI Regulation Enforcement hits three continents in one week. EU: transparency layer, market surveillance, high-risk AI systems obligations (Art. 13-15 GDPR-AI), non-compliance fines up to 6% global turnover. China: Companion-AI enforcement, persona shutdowns, ¥9,800,000 fine for emotional manipulation violation of CAC rules. Washington: voluntary frontier-model review framework, early-access approval required, access pending

AI · 08/17/2026

The AI Enforcement Era Landed on Three Continents at Once, and Only Washington Pretends It Didn’t

AI regulation enforcement arrived on three continents in one week: the EU AI Act's transparency layer went live, China issued its first companion-AI fines,…

9 MIN

READ →

Signal over noise.