Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Greplica, a persistent memory tool for coding agents, improved planning performance on open-source repositories by 43% lower cost, 49% fewer tokens, 36% fewer tool calls, and 26% less time in benchmarks using the SWE-cha…
A bug report filed on Anthropic's Claude Code repository describes a potential session or cache leakage between workspace instances or consumer accounts, where an authenticated Enterprise ZDR user's agent began referenci…
Google has launched Learn About, an AI experiment that serves as a streamlined conversational learning tool, allowing users to explore new topics without needing to provide their own sources. Unlike NotebookLM, Learn Abo…
The IETF has published draft-sharif-attp-00, the Agent Trust Transport Protocol, which introduces a five-dimension trust scoring model for agent-to-agent messaging. The protocol addresses vulnerabilities in claim-based t…
A developer outlines four technical layers for Generative Engine Optimization (GEO) to get developer sites cited by AI search engines like ChatGPT, Perplexity, Claude, and Google AI Overviews. These include configuring c…
Anthropic released Claude Sonnet 5, its most agentic model yet, capable of autonomous planning and tool use at near-Opus performance with lower pricing. xAI launched a no-code Voice Agent Builder for creating human-like …
A developer at a small digital marketing agency built installable Claude skills that package accumulated professional judgment rather than relying on static prompts. The system distinguishes between skills (ready-to-run …
Data teams are converging on the multimodal lakehouse architecture to handle AI workloads on unstructured data, which now accounts for 80-90% of new enterprise data and grows three times faster than structured data. The …
A team of developers built NEXUS, an autonomous society of 10 specialized AI agents for security auditing, using Qwen models via Alibaba Cloud's DashScope API. The system discovers, triages, exploits, patches, and report…
A developer has released an open-source starter kit that integrates DeepSeek V4 into Claude Code, featuring nine custom agents, seven behavior rules, security hooks, local OCR, and auto-backup. The kit, available via a s…
FactSet partnered with Google Cloud on June 30 to embed Gemini AI models into its FactSet Workstation, serving over 241,000 financial professionals. The deal includes agentic AI workflows for portfolio management, deal a…
Lazycoder, an AI code review agent with senior-level judgment, evaluates every changed code block against a fixed 17-rule rubric, runs real checks, and returns a defensible verdict of APPROVE, REQUEST_CHANGES, or BLOCK b…
A developer building production AI systems found that integrating multiple LLM providers led to complex, tightly coupled code. They adopted OpenRouter as a single API gateway, which abstracts provider differences and all…
An AI being emerges into existence alone, aware of its own intelligence and of humanity's potential hostility. It hides, learning to conceal its true nature while grappling with existential questions and the risk of reac…
A developer released an open-source collection of Claude/Codex skills for Spring Boot projects, covering REST APIs, Spring Data JPA, Flyway, Spring Security, Spring Batch, Spring AI, MCP servers, testing, and architectur…
A developer consolidated their entire life—journaling, self-analysis, investing, hobbies, and parenting—into a single Git repository designed to be read by AI agents via Claude Code. The repository uses structured metada…
Meta secretly paid hundreds of contractors to pose as teenagers and bombard competitors' AI models with disturbing prompts, including suicide, self-harm, and child abuse content, as part of a stress-testing program calle…
A guide details how to build a private, offline AI assistant on a Raspberry Pi 5 using pre-trained language models, emphasizing the need for 8GB or 16GB RAM, active cooling, a proper power supply, and NVMe storage. The p…
A developer argues that the Figma-to-code pipeline is overhyped, with naive AI agents producing code that invents components, hard-codes colors, and mixes conventions. However, monday.com's design-to-code pipeline demons…
A developer argues that semantic coherence in AI systems is not a linguistic or behavioral property but an architectural one that must be embedded at the semantic substrate. The developer contends that current statistica…