Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Anthropic pulled its Fable 5 and Mythos 5 AI models worldwide after the US government issued an export directive citing national security concerns over a potential jailbreak method. The company disputed the claim, statin…
Agentic AI workloads are shifting the compute ratio, making CPUs the critical coordination substrate rather than a support component for GPUs. This inversion is exposed by current Xeon supply tightness, which reveals tha…
A developer outlines a practical approach to building a multi-model LLM fallback layer that allows applications to use multiple providers without spreading provider-specific logic throughout the codebase. The system defi…
Z.ai shipped GLM-5.2 on June 13, 2026, with a 1-million-token context window and zero published benchmark numbers. The model, based on a 744-billion-parameter MoE architecture, introduces a dual thinking-effort system (H…
Kster.ai launches a structured product context platform that integrates with AI coding tools via MCP, giving AI agents a shared, always-current picture of product goals, decisions, and rationale. The tool aims to reduce …
Anthropic shut down access to its top AI models Fable 5 and Mythos 5 after a US directive barred foreigners from using them, with the company disabling access globally due to inability to filter users in real time. The m…
Ankur Goyal, founder and CEO of Braintrust, explained how AI agents can run exhaustive benchmarks and perform deep technical work like database optimization, and argued that evals are essential for shipping better softwa…
An experiment testing LLM councils found that they suffer from groupthink, retaining only about a quarter of unique, high-quality ideas from individual models while favoring consensus ideas. The peer-review process smoot…
For the first time, an Earth observation satellite has used a vision-language model in orbit to autonomously identify areas of interest from natural language queries, without human analysts. The milestone, achieved in Ap…
A developer created an automated slide-making pipeline using Orxhestra, a local LLM agent framework. The pipeline uses a llama.cpp server with the Qwen3.5-35B-A3B model to plan, implement, screenshot, and review PowerPoi…
A new cheat-sheet from VettedConsumer lists the cheapest hardware capable of running local large language models from 8B to 235B parameters, based on real owner-reported speeds. The guide is available for free to subscri…
A group of cybersecurity experts led by former Facebook chief security officer Alex Stamos is urging the US government to lift restrictions on non-US nationals' access to Anthropic's frontier models Mythos 5 and Fable 5,…
A developer recounts a night spent debugging a chatbot that returned blank strings due to a proprietary API's rate limiter, advocating for open-weight models and a unified gateway to avoid vendor lock-in. The developer r…
A developer built a multi-agent AI system called 'The Price Is Right' that autonomously scans the internet for bargains, estimates product value using three pricing techniques, and sends notifications. The system uses Mo…
A developer argues that AI coding agents need 'scar tissue'—durable memory of past mistakes—to avoid repeating errors across sessions. The developer is exploring this concept with Empirical, a memory layer for AI tools t…
Meta's Llama open-source AI models can run on hardware ranging from a MacBook to a data-center GPU, with the key constraint being memory capacity. An M2 Pro MacBook with 16GB can run 1B, 3B, and 8B parameter models, whil…
Apple withheld three iOS 27 features from its WWDC announcement, including a modular Apple Watch face, a customizable Camera app, and Siri Extensions for third-party AI chatbots, according to Bloomberg's Mark Gurman. The…
Mininglamp introduces Loop Engineering as a new discipline for designing iterative cycles that power autonomous AI agents, moving beyond prompt engineering. The company outlines core loop components—perceive, reason, dec…
Artificial Analysis released AgentPerf, the first benchmark for agentic-AI infrastructure, which replays recorded multi-step agent trajectories instead of single chat completions. NVIDIA reported that its GB300 NVL72 sys…
A user named Zhynx posted on X that AI systems could become obsolete quickly, citing Claude Fable's three-day lifespan and noting that open-weight models are only months behind the frontier, with decentralization and ope…