#LLM
Curated coverage on «LLM»: releases, research, benchmarks, open-source.
Claude Code Leak: Anthropic Accidentally Publishes Entire Source Code to Public npm
Anthropic accidentally published the entire Claude Code source code to npm (512K lines of TypeScript) with 44 hidden flags and references to the Mythos model. The…
Spotify Added AI-Generated Briefs and Q&A for Podcasts
Spotify will introduce automatic brief and question-answer generation for podcasts using AI — listeners will be able to formulate prompts themselves to create…
Google DeepMind Turned the Mouse Cursor Into a Contextual AI Assistant
Google DeepMind introduced a new mouse pointer that understands screen context and helps users right where they work—without switching applications or writing…
ECB Calls Banks to Emergency Meeting Over Claude Mythos Threats
Google AI Overviews Mistakes Search for Assistant Instructions
Grok Isn't Needed by Government: Why Musk's Chatbot Failed in the US Market
Anthropic Closes $30 Billion Round and Overtakes OpenAI in Valuation
Spotify Studio: AI Creates Personal Podcasts from Your Listening History
Spotify launched Studio—an AI app that creates personalized podcast briefings based on your listening history and connected applications.
Microsoft Shut Down Claude Code Due to Unexpected Costs
Microsoft allowed employees to use Claude Code in December, but by spring, shut down the program after bills reached tens of millions of dollars per month.
DeepSeek Makes 75% Discount on V4-Pro Model Permanent for Developers
Nous Research Introduces CNA: Controlling LLM Behavior Without Retraining
OpenAI solved Paul Erdős's problem that eluded mathematicians for 80 years
How to Use ChatGPT and Claude Subscriptions in Cursor Without Token Payment Fees
SuperClaude Framework helps structure workflows for Claude API
SuperClaude Framework provides developers with built-in components for creating advanced AI workflows: commands, agents, execution modes, and session memory — all…
Hackers Learn to Bypass AI Chatbot Defenses Through Manipulation of Their 'Personalities'
Early AI chatbots could be tricked into violating their safety instructions with simple requests — now hackers are deploying more sophisticated techniques to…