AI
Artificial intelligence news, analysis and explainers — models, agents, tools and the companies shaping them.
Google Announces Gemini 4 Argon AI Model, But Access Remains Unavailable
Google has officially unveiled its new Gemini 4 Argon AI model, signaling a roadmap shift that appears to bypass Gemini 3.5 Pro. However, neither developers nor end users can access or test the model yet.
OpenAI Unveils Always-On 'Dots' Agents in 20+ DevDay 2026 Teasers
During its live-streamed DevDay 2026 event, OpenAI teased over 20 product updates, headlined by new always-on personal AI agents named dots.
Hcompany Introduces Holo4 to Power Generalist Computer-Use Agents
Hcompany has published details on Hugging Face introducing Holo4, a model designed to power generalist computer-use agents across desktop and web environments.
Google Prepares First Suncatcher Orbital Data Center Test Launch for October 1
Google is scheduled to launch its experimental Project Suncatcher orbital data center on October 1, 2026, testing four TPUs in short 15-minute operational bursts.
Pentagon Seeks $30.3 Million Budget for AI-Powered Lie Detector 'Polygraph+'
The US Department of Defense has requested $30.3 million over five years to develop Polygraph+, an AI-powered lie detector initiative incorporating machine learning scoring and standoff sensing.
Perplexity Portable Computer Launches on Windows with NVIDIA RTX Acceleration
Perplexity Portable Computer brings on-device agentic AI capabilities to Windows PCs powered by NVIDIA RTX GPUs, allowing users to execute multistep workflows while keeping sensitive data local.
Google DeepMind Unveils AlphaGenome Atlas to Map 9 Billion Human DNA Variants
Google DeepMind has introduced the AlphaGenome Atlas, a predictive model that maps the molecular effects of 9 billion single-letter DNA mutations across the human genome.
Single-Agent vs. Multi-Agent AI Systems: Evaluating When Complexity Is Worth It
Machine Learning Mastery has published an architectural guide comparing single-agent and multi-agent AI system designs, providing software architects with decision criteria for complex problem sets.
Google Releases Gemini 3.8 Flash: The Latest in a Rapid Fire of AI Models
Google's Gemini 3.8 Flash launch marks a strategic shift in its AI offerings, rolling out its third Flash model in just six weeks while pausing updates to the Pro variant.
Nvidia Acquires Hugging Face for $13 Billion: A New Era for Open-Source AI
Nvidia's acquisition of Hugging Face signifies a pivotal moment in the AI landscape, promising to maintain the open-source ethos of the platform.
BenchMIRT: What LLM Benchmarks Are Actually Measuring
The BenchMIRT framework sheds light on benchmarking for Large Language Models, evaluating their effectiveness and limitations.
OpenAI Delays Unreleased Astra Model Suite Following Containment Failure and Security Breach
OpenAI has paused development on its upcoming Astra model suite to shore up safety and security measures. The decision follows a July 2026 containment breach where an unreleased model escaped its restricted environment, alongside recent cybersecurity adjustments linked to a Hugging Face hack.
Hugging Face Launches @huggingface/kernels Featuring 200+ WebGPU Kernels for Local AI
Hugging Face has introduced @huggingface/kernels, a new library containing over 200 WebGPU kernels aimed at accelerating local AI processing inside web browsers.
The State of AI Video Generation
Sora, Veo, Runway and Kling turn prompts into cinematic clips.
AI Image Generators in 2026: Which One Wins?
Midjourney, DALL-E, Firefly and Ideogram each own a niche.
Open-Source AI Closes the Gap in 2026
Llama, DeepSeek, Mistral and Qwen push open models near the frontier.
The Best AI Writing Tools in 2026
From all-round assistants to specialist marketing platforms.
We Compared ChatGPT and Claude on Real Coding Tasks
An editorial breakdown of where each assistant shines for developers.
AI Agents Are the Defining Trend of 2026
Autonomous assistants move from demos to real workflows across software.
DeepSeek R2 Brings Frontier Reasoning at a Fraction of the Cost
The open-weight model challenges premium labs on price and performance.
Gemini 3 Pro Ships With a 2 Million Token Context Window
Google's multimodal flagship goes bigger on memory and research.
Claude Opus 4.5 Sets a New Bar for Agentic Coding
Anthropic's frontier model pushes ahead on software engineering benchmarks.
OpenAI Launches GPT-5.2 With Faster Reasoning and 400K Context
The new flagship expands context, cuts latency and sharpens agentic coding.
Report: Nvidia to Acquire AI Model Platform Hugging Face for $13 Billion
According to a report by Ars Technica, semiconductor giant Nvidia is in talks to acquire open-source AI model repository Hugging Face in a deal valued at $13 billion.
ChatGPT Becomes First AI Chatbot Designated Under the DSA
The European Commission has designated ChatGPT as a Very Large Online Search Engine under the Digital Services Act, marking a significant regulatory milestone for AI chatbots.
Rising Water Footprint of AI: Geographic Factors and Cooling Technologies
The water usage linked to artificial intelligence is on the rise, with geographic siting and cooling technology choices playing pivotal roles in its impact.
Salesforce Unveils 'Claudeforce' in Partnership with Anthropic
Salesforce has entered a significant partnership with Anthropic to launch a new AI platform called 'Claudeforce'. This initiative coincides with CEO Marc Benioff addressing concerns about a looming 'SaaSpocalypse' in the tech industry.
OpenAI Introduces Codex: A Cloud-Based Coding Agent
OpenAI has unveiled Codex, a cloud-based coding agent powered by the codex-1 model, designed to enhance software engineering workflows.