Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
The New Mexico Supreme Court fined Santa Fe defense attorney Stephen Aarons $5,000 and held him in contempt on Wednesday for a criminal appeal brief that contained fabricated witness testimony and false statements genera…
A community-maintained handbook compiles ChatGPT slash commands, UI shortcuts, Canvas triggers, and custom command engines into a single reference, noting that OpenAI publishes no official master list. The guide categori…
LLMjacking has grown from a proof-of-concept threat documented by Sysdig in May 2024 into a commercialized criminal enterprise, according to a March 15, 2026 research note from the Cloud Security Alliance AI Safety Initi…
A developer spent three months testing AI automation frameworks and prompt collections, concluding that most are recycled listicles with no testing behind them. The developer reports that three workflow patterns — idea-t…
Cognition launched Devin Fusion, a coding agent architecture that pairs a frontier "lead" model for planning with a cheaper "sidekick" model for execution, delivering up to 39% better efficiency and a 46% cost reduction …
A user testing a multi-GPU LLM development machine reported that the system held up well under sustained load running the Qwen3.8 27B model, despite the machine shipping from 45 Drives with the wrong CPU and only 8 cores…
AI-generated code most often fails because models hallucinate library versions, invent nonexistent function parameters, or lose context in long conversations, according to a debugging guide that recommends pasting exact …
The New Mexico Supreme Court held attorney Stephen Aarons in direct contempt of court for filing a brief containing "false testimony from wholly fabricated witnesses" generated by ChatGPT, fined him $5,000 payable to the…
A developer has open-sourced AgentJIT, a just-in-time trajectory compiler that traces an AI agent's dynamic tool chains and compiles them into deterministic Python AST pipelines, cutting execution time from roughly 30 se…
Anthropic said in a threat intelligence report on Thursday that several actors used its Claude AI models for weapons development, cyber operations, surveillance and fraud, including a China-based actor that modeled a Tai…
Baidu posted a Forward Deployed Engineer Intern opening at its Mountain View, CA office, an on-site, full-time role lasting 3 months with the possibility of extension. The intern will help customers integrate Baidu's Que…
Elon Musk said on September 11 that xAI has delayed Grok 4.7, its next-generation AI model, for "a few more days" of reinforcement-learning tuning, pushing back a release he had previously pegged four separate times sinc…
Anthropic said users of its Claude chatbot in Houthi-controlled northern Yemen attempted to develop advanced missiles, including a multi-variant hypersonic-glide missile and a warhead using mobile phone hardware for mid-…
AWS introduced a production monitoring approach combining Amazon Bedrock AgentCore Evaluations and AWS DevOps Agent to track multi-agent systems, demonstrated on a four-agent airline reservation system. AgentCore Evaluat…
A Hacker News user asked how successive generations of AI models are built, questioning whether Anthropic's Opus is trained from scratch, derived from Sonnet, or largely the same model with more compute and training. The…
A developer used Anthropic's Claude to generate a 3D interactive map and 64 documentation sheets for a Pulumi-managed homelab, publishing the result at homelab.robbeverhelst.com. After Claude's first attempt to render th…
OpenAI deployed roughly 10,000 AI agents simultaneously in August 2026 and solved the Navier-Stokes Millennium Prize Problem in 88 hours, according to The Verge. The result highlights a broader shift in AI agent architec…
DeepSeek launched V4.1-Flash on Thursday, a 552-billion-parameter mixture-of-experts model that activates roughly 8 billion parameters per input token and 16 billion per output token, with a one-million-token context win…
A new Python library called litelm extracts the core routing and message-translation functionality of litellm into roughly 2,900 lines of code with just 2 dependencies, openai and httpx, dropping the proxy server, cachin…
GPT-6 Astra accepts text and images but rejects mp4 video input through its API, so developers are working around the limitation by sending sequences of timestamped frames as input_image parts alongside timestamps in the…