Search Latest News

Microsoft Research's Mirage video world model stores scene information in latent space rather than pixel-based point clouds, significantly reducing compute requirements. This efficiency gain accelerates the practical deployment of persistent spatial memory in video generation systems used across entertainment, simulation, and autonomous systems.

Google Cloud introduced Open Knowledge Format (OKF), a standardized specification that converts scattered organizational documents into portable Markdown files with YAML metadata for AI agent consumption. This standardization reduces friction in knowledge integration pipelines, enabling faster AI deployment across enterprise systems.

The Ugreen FineTrack 2 Bluetooth tracker offers extended battery life and unique features compared to Apple's AirTag. Consumer preference for longer-lasting tracking hardware reflects broader demand for low-maintenance IoT devices suitable for supply chain asset monitoring.

KPMG published a report on AI in business that contained fabricated case studies involving UBS, the NHS, and other organizations. GPTZero CEO Edward Tian, who helped uncover the errors, warns of "secondary hallucinations," flawed claims from trusted consulting firms that spread unchecked. KPMG has since pulled the report. The article KPMG fabricated AI case studies in a report designed to sell clients on AI adoption appeared first on The Decoder.

AI coding agents like Claude Code or Codex reliably find the right file but miss most of the critical lines within it. The new SWE-Explore benchmark is the first to test code search separately from the actual repair, and it shows that without enough context, even the best fix will fail. The article AI coding agents find the right file but miss the exact lines that matter, study shows appeared first on The Decoder.

Amazon CEO Andy Jassy and executives from five other tech companies reportedly alerted the Trump administration to security vulnerabilities in Anthropic's Fable model, prompting a White House export control order that forced the model offline within hours. This incident demonstrates how investor concerns can rapidly escalate to government intervention in AI model deployment, creating regulatory uncertainty for AI companies.

A new AI model called "Count Anything" has been developed to count objects in any type of image using only text prompts, from crowd photos to microscope cell samples. This capability addresses a previously difficult computer vision task with broad applications across healthcare, manufacturing, and logistics industries.

Microsoft CEO Satya Nadella warned against "token-maxing"—deploying frontier AI models indiscriminately—while admitting he himself engages in the practice because it is "addictive." This tension between cost efficiency and model capability reflects an industry-wide challenge in optimizing AI deployment economics as compute costs escalate.

Google Research's Gemini-SQL2 achieved 80.04 percent accuracy on the BIRD text-to-SQL benchmark, significantly outperforming OpenAI and Anthropic models. This advancement in natural language-to-database query translation enhances enterprise data access and analytics capabilities.

Microsoft and three Chinese universities have developed SkillOpt, a method that optimizes instruction documents for AI agents using principles from traditional model training. A simple Markdown file is enough to boost GPT-5.5 by about 23 points on procedural tasks, and the same file transfers across models and agent environments like Codex and Claude Code. The article Microsoft's SkillOpt boosts GPT-5.5 by using nothing but a trained Markdown file appeared first on The Decoder.

Anthropic's Claude Fable 5 hits 88 percent accuracy on the hardest FrontierMath tier, a massive jump from Opus 4.5, which sat below 10 percent in early 2026. OpenAI's GPT-5.5 reaches about 75 percent on the same tier. The pace of improvement in AI math keeps accelerating. The article Claude Fable 5 outpaces GPT-5.5 by 13 points on FrontierMath's toughest problems appeared first on The Decoder.

Meta's internal AI infrastructure costs have reached billions annually, prompting the company to implement token-usage governance starting in 2027 through a system called 'AI Gateway.' This signals that large-scale AI deployment economics are forcing technology leaders to shift from unconstrained capacity expansion to managed resource allocation.

This item is a product review guide for VPN routers ranked by speed, security, and reliability with no quantifiable data or market impact information. It has no relevance to critical minerals, AI policy, or technology sector investment analysis.

Visa and OpenAI have established a partnership enabling AI-prompted financial transactions, advancing automated purchasing capabilities through large language models. This integration of AI decision-making into payment processing raises questions about transaction verification and liability frameworks.

OpenAI now lets Codex users bank their rate-limit resets and trigger them manually instead of watching them expire on a fixed schedule. If you hit your usage cap mid-session, you can cash in a saved reset right away instead of waiting. Users on the Go, Plus, Pro, and Business plans each get one free reset to start. Plus and Pro users can also invite friends to unlock extra resets. The article OpenAI kicks off the AI price wars with flexible rate-limit resets for its Codex coding agent appeared first on The Decoder.

Claude Fable 5 tops the Artificial Analysis Intelligence Index with 64.9 points and sets records in five of ten benchmarks. But the gain over Opus 4.8 is just 5.7 percent at double the token price. Safety filters with fallback routing push costs even higher. The article Anthropic's Claude Fable 5 costs twice as much for 5.7 percent more performance appeared first on The Decoder.