Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Nivedita Mukherjee published an architectural analysis comparing proprietary generative AI abstraction, exemplified by IBM watsonx Code Assistant for Z, against open DevOps orchestration using Zowe CLI, Endevor Web Servi…
Anthropic said on September 10 that it disrupted several allegedly malicious uses of its Claude models over the past eight months, including a Russia-linked cyber espionage campaign and what it called the largest "illici…
GPT-Live-1, a new full-duplex voice API, is now available to developers at $0.05 per minute, allowing the model to listen and speak simultaneously and detect user interruptions in real time. The model scored 80.1% on int…
Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co. Ltd. released DeepSeek-V4.1-Flash, a 552-billion-parameter mixture-of-experts model it says beats its larger DeepSeek-V4-Pro on performance, cost, s…
A user detailed a home AI inference rig built on an AMD Epyc Rome 7532 CPU with 128GB DDR4, roughly 5TB of NVMe storage, and 4x AMD V620 GPUs on an open-air crypto-mining bench. The rig runs a LiteLLM frontend to vLLM se…
NASA's Marshall Space Flight Center Impact AI team and IBM Research released INDUS-SDE, a specialized language model pre-trained on NASA's Science Discovery Engine data that reached 78.1% top-1 masked language modeling a…
A refurbished Miniforum UM790 Pro with 64GB RAM and 1TB NVMe, powered by an AMD Ryzen 9 7940HS CPU with Radeon 780M GPU, was purchased for $537 and used to run local LLMs via Ollama and Open Web UI on OpenBSD and Linux. …
A head-to-head test of Cursor, Windsurf, and Claude Code on a 40k-line Next.js project found Cursor's Composer needed 4 tries to fix a React useEffect race condition, Windsurf's Flow took 2 attempts, and Claude Code was …
DeepSeek released V4.1 Flash with a 1M context window, FP4 KV cache, and cross-layer attention reuse, while Alibaba open-weighted Qwen3.8-2.4T-A95B, a Qwen-Max-class mixture-of-experts model with 2.4 trillion parameters …
A new blog post argues that the core primitive for working with AI is shifting from the LLM API to the Harness API, a model the author calls Harness as a Service (HaaS). The post names Anthropic's Claude Code SDK as the …
OpenAI has signaled a possible "Data agent for ChatGPT Work" but has not published customer-facing documentation for a standalone feature by that name. The company's July 9, 2026 ChatGPT Work announcement describes a gen…
An analysis of Anthropic's Claude Developer Platform found that its tool catalog and memory store handle retrieval differently: the tool search API offers regex- and BM25-based ranked search that returns up to five match…
Security researchers demonstrated an attack dubbed EvilFont in which a Microsoft Word 97-2003 document (out.doc) built with the open-source EvilFontTool displays the harmless sentence "This is a test file, nothing import…
A developer's guide outlines the technical steps for getting a site cited in Google AI Overviews, the AI-generated summaries powered by Gemini that appear above traditional search results. The post argues that crawlabili…
OpenRouter's tracking of the z-ai/glm-4.7-flash endpoint at venice/fp8 recorded a B3IT total-variation level shift of 0.38, with the endpoint now in active monitoring status. The endpoint does not return logprobs, making…
Google researchers Zhongyi Zhou and Ruofei Du introduced ToolGrad, a data generation framework that reverses the traditional tool-use dataset paradigm by generating ground-truth tool-use chains before user queries, prese…
Anthropic PBC's latest threat intelligence report, covering Claude misuse it disrupted between December and August, says AI now lets individual hackers run campaigns that once required state-backed teams, with breaches c…
OpenAI withdrew its sponsorship of Caltech's AI math hackathon on Thursday after current and former Caltech mathematicians published an open letter on Wednesday warning the event "is likely to have destructive impacts fo…
Redis launched LangCache, a fully managed semantic caching service now in public preview on Redis Cloud, which Redis reports can cut LLM API costs by up to 90% and return cache-hit responses up to 15x faster than re-quer…
A former B2B SaaS web developer who switched to inference engineering about a month ago published a skills map arguing the field requires only three service categories: routing and scheduling via Nvidia Dynamo or llm-d, …