A developer demonstrates using DeepEval to evaluate an enterprise agent, contrasting its test-case-first paradigm with RAGAS's batch evaluation. The implementation uses a custom judge LLM (glm-4-flash) and reveals that l…
Nearly 200 venture-backed startups, organized by the Little Tech Association, sent letters to President Trump and other officials on July 22 urging the U.S. government not to restrict access to Chinese open-weight AI mod…
Data & Society released a handbook on using government procurement to reduce the environmental impact of AI, based on 10 months of fieldwork and five months of co-design with city staff and green software practitioners. …
A developer built facet, a Go CLI tool that creates disposable multi-repo workspaces for Claude Code agents, solving the problem of multiple agents fighting over a single dirty git index. The tool reads GitHub issues to …
A developer explains Retrieval-Augmented Generation (RAG), a technique that gives LLMs access to actual data before answering, reducing hallucinations. The pipeline involves indexing documents into a vector store during …
A developer connected three MCP servers to a single AI agent and discovered that while MCP standardizes tool communication, it lacks governance for multi-server setups. The developer built Fabric, an open-source control …
Google split its eighth-generation TPU into two separate chips — TPU 8t for training and TPU 8i for inference — at Cloud Next 2026, driven by the demands of AI agents that generate 50 to 100 times more inference transact…
The White House is finalizing a voluntary 30-day pre-release AI review framework with OpenAI, Anthropic, and Google, targeting completion before August 1, that gives federal agencies 30 days to evaluate frontier models b…
Microsoft integrated Anthropic's Claude Opus 5 into Microsoft 365 Copilot on July 25, making it available in Word, Excel, PowerPoint, Copilot Chat, Cowork, and Studio alongside OpenAI's GPT-5.6. The integration marks the…
Moonshot AI will release the full open weights of its Kimi K3 model on Sunday, July 27 — a 2.8-trillion-parameter Mixture-of-Experts model with a 1-million-token context window and native vision that topped Arena.ai Fron…
Nvidia and South Korea's SK Group announced a $500 billion-plus strategic partnership on July 24 covering a 2-gigawatt AI factory in South Korea and long-term HBM4 memory co-development with SK Hynix, locking in Nvidia's…
SpaceX's S-1 IPO filing reveals Anthropic pays SpaceXAI $1.25 billion per month for exclusive access to all 220,000 Nvidia GPUs at the Colossus 1 facility in Tennessee through May 2029, the largest AI compute deal in his…
Shopify is launching a new storefront theme with 93% fewer lines of code than its current Horizon theme, driven by the need to make code readable for AI agents like its Shopify Sidekick assistant. Product director Ben Se…
A fallen power line in Northern Virginia exposed how AI data centers struggle to respond to grid disruptions, according to a TechCrunch analysis. The incident highlights the need for better backup systems and grid coordi…
A technical analysis reveals that large language models (LLMs) have a universal refusal mechanism embedded in their weights, but safety guardrails are inconsistently activated for non-English inputs because most alignmen…
The AI industry is shifting its focus from raw model intelligence to 'intelligence per dollar,' as many models become good enough for business tasks and buyers prioritize cost efficiency. Amazon's rebuild of Alexa+ exemp…
Meta is testing an AI app called StoryKit that generates personalized bedtime stories for children, allowing parents to create characters, settings, and moral lessons in seconds. The app, currently in pilot countries, ai…
Cornerstone University, a Christian university in Grand Rapids, Michigan, has formed a President’s Artificial Intelligence Advisory Board to develop a coherent AI strategy and curriculum, aiming to graduate students who …
The Guardian reports that the AI jobs apocalypse is unlikely to arrive soon, citing Anthropic's analysis of AI's impact on employment. Anthropic co-founder Dario Amodei had previously claimed AI could eliminate half of a…
Anthropic released Fable-lite, a smaller version of its AI model, while Block introduced Buzz, a new product, according to Tom Krazit's Runtime newsletter. Google unveiled a new security model, Poolside released a new op…