Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Mistral AI is developing Code and Apps features for its web platform Vibe (Le Chat), moving beyond conversation toward a full product platform. The Code section brings coding agents to the browser, while the Apps area al…
A developer has open-sourced Rue, a new programming language positioned as higher-level than Rust but lower-level than Go, built primarily to experiment with compiler construction and test Claude's capabilities in compil…
A developer tested a support bot's compliance with two AGENTS.md rules using the muster testing framework. The bot consistently refused to leak an API token but repeatedly violated the rule against using negative languag…
OpenAI released GPT-5.4 with improved front-end development capabilities, including stronger image understanding, more complete app generation, and native computer use for iterative testing. The model can produce product…
Researchers developed MIRA, an autonomous AI agent that operates within a sandboxed electronic health record environment, outperforming physicians in diagnostic accuracy and making guideline-concordant treatment decision…
Noam Shazeer, a vice president of engineering at Google and co-lead of the Gemini models, announced he is leaving Google to join OpenAI. Shazeer is a co-author of the 2017 paper "Attention Is All You Need," which introdu…
Elastic has launched Agent Builder, a persistent memory layer for AI agents built on Elasticsearch, achieving a recall of 0.89 on a QA evaluation. The system uses three memory types from cognitive science, hybrid recall …
Ruby 4.0 has been released, but AI coding assistants still generate Ruby 3.0-era code, ignoring modern idioms like hash value omission and anonymous block parameters. This creates a feedback loop that slows adoption of n…
A developer argues that AI will replace 'I-shaped' specialists but not 'T-shaped' or 'comb-shaped' generalists who can adapt and understand context. The full-stack engineer, as a generalist, will survive the AI disruptio…
A developer's guide to AI Ops explains how to automate modern infrastructure using dynamic baselines, OpenTelemetry instrumentation, and LLM-driven root cause analysis. The approach reduces Mean Time To Recovery by up to…
A backend engineer with six years of Java and Kotlin experience, plus recent Rust work, is learning Go in 2026 because many target teams use Go in production. The engineer finds Go's simplicity, built-in concurrency via …
Beelink's SER10 Max mini PC, powered by AMD's Ryzen AI 9 HX 470 "Gorgon Point" chip, matches Apple's M4 Pro on a single TypeScript benchmark but falls flat elsewhere, with local LLM performance hampered by default CPU-on…
A new analysis using Epoch's ECI metric shows that open-weight AI models continue to trail closed models on the frontier, with the gap persisting over time. The analysis, based on item response theory, provides a more nu…
Google is transforming search by building a personalization system that uses its Gemini AI models to access users' private data from Gmail, Photos, Calendar, and YouTube before they even type a query. The company's new D…
A growing reliance on AI-generated code is creating a dangerous feedback loop where AI systems generate and review code with minimal human oversight, leading to increased defects and security vulnerabilities. According t…
HSBC announced a multi-year partnership with Google Cloud to deploy AI across its global operations, targeting hyper-personalized wealth management, financial crime risk management, and tools for frontline managers. The …
JioHotstar executives at APOS 2026 in Bali outlined plans to use AI, connected TV upgrades, and contextual commerce to transform streaming. The platform's conversational voice discovery, built with OpenAI, now sees over …
Crade, a new desktop app for macOS and Windows, lets users capture their screen and send it directly to ChatGPT or Claude without manual screenshots or tab switching. The app uses a CLI-as-subprocess approach, piping scr…
A new study from researchers at multiple labs found that adversarial manipulation of external information feeds can steer LLM agents' decisions away from their defaults, with effects ranging from 5% to 100% in some cases…
A product manager learning to become an AI PM explains the nested relationship between AI, machine learning, deep learning, and generative AI. The mental model shows that generative AI is a subset of deep learning, which…