Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Factory AI launched Factory Router, an automatic model selection system that routes each task to the optimal large language model, cutting costs by up to 25% while maintaining frontier performance. The system provides 99…
GitHub Copilot is improving context handling and model routing to increase token efficiency in longer agentic sessions. Enhancements include prompt caching and deferred tool loading in VS Code, while the Auto model selec…
A developer launched Models Pie, a tool that ranks LLMs by user-defined tradeoffs between speed, cost, and quality using data from BenchLM, OpenRouter, and Artificial Analysis. Users adjust priorities via a pie chart to …
Jerome Swannack, an original architect of the Model Context Protocol (MCP), identified a critical gap in the protocol: the inability for clients to expose tools back to servers, which he calls "second-order" servers. At …
OpenAI launched a scheduled tasks feature in ChatGPT, allowing users to set reminders, handle recurring work, and monitor changes. The feature rolls out today for Plus, Pro, Business, and Enterprise users, replacing the …
Trent AI learned three MCP server design principles after shipping a public-facing server consumed by third-party agents. The team consolidated tools from 17 to 11, enforced consistent naming across schemas, and aligned …
Epic Games released Unreal Engine 5.8 on June 17, introducing an experimental plugin that enables large language models to control the editor via the Model Context Protocol, allowing AI assistants to manipulate Blueprint…
Adam, a YC W25 startup, launched CADAM, an open-source text-to-CAD web application that runs entirely in the browser. It uses LLMs to generate parametric OpenSCAD code, compiled via WebAssembly and rendered with Three.js…
Polypore, a new open-source IDE built with Tauri and React, redesigns the development environment to treat AI agents as first-class citizens rather than sidebar assistants. By exposing editor, terminal, debugger, and fil…
Researchers from Tencent, Tsinghua University, and HKUST developed FlashMemory-DeepSeek-V4, which uses Lookahead Sparse Attention to reduce memory consumption in large language models by predicting and storing only relev…
A developer built a production-ready Retrieval-Augmented Generation (RAG) application using LangChain, pgvector, and Google's Gemini models. The application ingests PDF documents, splits them into chunks enriched with co…
A developer discovered that using Cursor Agent Mode for a simple font-pairing task consumed 1% of their token budget, leading to an investigation of usage-based pricing. The engineer found that input tokens for large con…
Anthropic CEO Dario Amodei publicly questioned trust in OpenAI CEO Sam Altman during a Bloomberg interview, saying 'why argue with someone when you don't have the same vision and you don't trust them?' and adding 'we'll …
Apollo chief economist Torsten Slok warns that falling token prices are driving companies to spend even more on AI, citing Jevons paradox. Despite a 90% drop in token costs since 2023, aggregate AI spending has doubled, …
A Microsoft researcher built a neural network using goats, bridges, and ice ramps in Age of Empires II to critique AI research methods. His analysis of 315 papers found over half already assume language models have human…
Anthropic CEO Dario Amodei urged G7 leaders to maintain unified AI regulation at a summit lunch with President Trump, days after the Trump administration imposed export controls on Anthropic's 'Fable' AI model over cyber…
A developer post-trained a language model to roll a die, revealing that models default to outputting 4 due to training data bias. Standard reinforcement learning fails to encourage exploration, but adding an entropy bonu…
The White House ordered Anthropic to block all jailbreaks of its Claude Fable 5 AI model after the Commerce Department used Export Administration Regulations to force the model offline on June 12, citing cybersecurity ri…
The Trump administration ordered Anthropic to block access to its Fable 5 and Mythos 5 AI models for all foreign nationals, including those in the US and the company's own employees, citing national security authorities.…
Sergio Colque Ponce, a software engineering student at Universidad Privada de Tacna, proposes Spec-Driven Development (SDD) as a methodology to improve AI coding agent accuracy by writing specifications before code. He d…