Full-text search across 68681 articles. Combine with topic and date filters; results sorted by relevance.
DesignArena, the AI evaluation startup founded by Grace Li and Kamryn Ohly, launched a video arena in which two models receive the same prompt and users compare the short films each produces by writing HTML, CSS and Java…
Two independent projects, ZacOS 9 and Mavernix, have emerged to bring vintage Macintosh-style desktop environments to Linux, with ZacOS 9 built on Wayland and acknowledging Claude and Copilot LLM coding assistance on its…
Google's A2UI project released swappable inference formats for its agent-optimized generative UI protocol, with the new A2UI Express format cutting latency by up to 3x and using up to 3x fewer tokens than Direct JSON, wh…
Researchers submitted ACARe (Agent-Context-Aware Reranker), a compact 4B LLM trained to rerank documents using the agent's broader task and reasoning context rather than the sub-query alone, to arXiv on 26 October 2026. …
Axiom Math released the first complete computer-checked proof of the Classification of Finite Simple Groups, a result originally spanning roughly 15,000 pages across more than 500 articles by about 100 mathematicians. Ov…
The Agentic AI Foundation presented its Outstanding Technical Contribution Award to an MCP Python SDK maintainer at AGNTCon + MCPCon North America on the second day of the event, recognizing work that included shipping t…
A Spidra tutorial demonstrates converting the arXiv paper "Attention Is All You Need" (identifier 1706.03762, version 7) from PDF to Markdown using Python 3.9 or later, the Spidra Python SDK and the arxiv package. The co…
Hands-on testing of OpenAI's three-model GPT-6.1 family found that mid-tier GPT-6.1 Terra costs roughly half of flagship GPT-6.1 Soul and runs faster, but produced results closer to budget-tier GPT-6.1 Luna on complex ta…
A developer built shell-ghost, an open-source zsh hook that uses preexec and precmd to capture each command and its exit code, then sends them to a local Gemma 3 4B model via Ollama's HTTP API to print a single eerie one…
GitHub made assisted permissions generally available in GitHub Copilot CLI, the GitHub Copilot app, and Visual Studio Code, letting a large language model evaluate each request to run a command, access a file system path…
The Agentic AI Foundation added four projects at AGNTCon + MCPCon North America 2026 in San Jose, growing from three founding projects to ten in its first year, executive director Mazin Gilbert announced. Alibaba contrib…
OpenAI released GPT-6.1 Terra on Monday, October 26, with standard API prices 50% lower than GPT-5.6 Terra, and made it available to Plus, Pro, Enterprise and Edu users in ChatGPT Work and Codex and to developers as gpt-…
LiteLLM's postmortem attributes an October 21, 2026 supply chain attack to a master key committed to a public GitHub repository on May 27, 2026 and never revoked, leaving it active for 136 days until an attacker used it …
A developer built FableCut Studio, a browser-based open-source video editor that lets AI agents edit an existing timeline directly through the Model Context Protocol rather than generating a new video from a prompt. Stud…
Mythex, an AI app builder based in Dubai, released a remote MCP server at mcp.mythex.ai/mcp that gives coding agents in Claude, Cursor and Codex a cloud sandbox, live preview, Postgres database, file storage, secrets and…
A Google DeepMind paper with the National University of Singapore, "The Illusion of Stochasticity in LLMs" (arXiv 2604.06543, presented at COLM 2026), found large language models fail to produce outputs matching specifie…
A developer known as onaga00 has released ProjectStonehenge, a native x86_64 PC port of the PlayStation 2 game Ace Combat 04: Shattered Skies (NTSC-U, SLUS-20152) that was built by translating the original game code func…
Google Cloud introduced IAM Helper, a suite of agent-facing building blocks that lets autonomous agents run end-to-end Identity and Access Management workflows on Google Cloud IAM, including Policy Intelligence and Privi…
Warp bills agent work in AI credits that its documentation calls "a usage unit, not a token or a fixed number of prompts," with cost varying by model, input and output volume, caching, tool calls, summary steps and extra…
Reddit's internal AI prototyping platform, Reddit Proto, has produced about 700 prototypes and reached 100% adoption across the company's UX and product teams, according to Blair Metcalf, Reddit's principal UX engineer a…