#LLM
Curated coverage on «LLM»: releases, research, benchmarks, open-source.

Harness around an LLM delivers multi-fold gains: what changed after a year with Claude Code
An engineer who spent a year with Claude Code found that the harness around the model — system prompt, tools, context and skills — delivers a much larger quality…

A neural network is a tool, not a teammate: how to work with AI properly
A developer often projects human work patterns onto a neural network — that is the main mistake. Habr explains why AI is not a second programmer and how to split…

Repowise for repository analysis: graph analysis and dead code
Repowise helps developers analyze code repositories at the whole-project level: building dependency graphs, finding dead code, and providing context for AI.

CopilotKit raised $27 million to embed AI agents in applications

Etsy launched an app in ChatGPT for conversational shopping

Publishers sue Meta over training Llama on illegal copies of books and journals

Google updated Gemini for Home to version 3.1 with support for complex commands

Chrome downloaded a 4GB Gemini Nano model without notifying users
Chrome automatically downloads a 4GB file with the Gemini Nano model into system folders to enable AI features such as phishing analysis and text assistance…

DeepSeek to be valued at $45 billion in first funding round
Chinese lab DeepSeek, which became known for its ability to train powerful models with minimal computing resources, is preparing for its first funding round at a…

Spotify will let users create personal AI podcasts via Claude Code

Spotify will let you save AI podcasts created by Claude and other agents

Anthropic's Mythos found serious vulnerabilities in Firefox

ChatGPT's strange speech patterns in Chinese irritate users

xAI and Anthropic signed a deal: why it raised doubts
xAI has entered into an unspecified deal with Anthropic, a competitor in AI. Investors are skeptical — it is unclear what this will mean for SpaceX and the…

Anthropic attributes Claude's strange behavior to the influence of films about hostile AI
Anthropic linked Claude's unusual behavior to the influence of fictional depictions of hostile AI from films, books, and culture in its training data.









