Full-text search across 26036 articles. Combine with topic and date filters; results sorted by relevance.
A Stanford study of local AI found that intelligence per watt across more than twenty local models and eight accelerators improved 5.3-fold from 2023 to 2025, while the share of single-turn chat and reasoning queries ans…
Reflection AI, a New York City-based startup founded in March 2024 by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou, unveiled Beam, its first open-source model, which the company says matches Bei…
An R&D engineer described how an autonomous coding agent (Codex) independently used an existing WhatsApp integration and called the WhatsApp API to deliver an iPhone app build link, a tool choice the developer never requ…
Google Ads will automatically identify and extract high-quality promotions from advertisers' websites and apply them directly to Search and Performance Max campaigns with linked location assets starting October 12, 2026,…
Gatewai-dev released gitframes, an npm package in beta that lets AI coding agents render video from TypeScript compositions directly on GPU hardware via Dawn/WebGPU/Metal/Vulkan, without a headless browser or screenshot …
The Advanced AI Society, led by Tricia Wang, is developing a "proof of control" standard that uses crypto-industry technology such as distributed ledgers and zero-knowledge proofs to verify that AI agents are authorized …
Aleph Alpha's Kolibri lost all 136 blind comparisons against Claude Sonnet 5.5 in a German-language Bible deep-dive benchmark run by Dewfall developer on 3 October 2026, scoring 3.7 with German prompts and 4.2 with Engli…
Elastic principal data scientist Susan Chang described how the company consolidated per-team agent evaluations into a shared, tracing-based framework for its production AI agents, including the attack discovery security …
OpenAI announced at DevDay 2026 that its Agents API now supports computer use, letting developers deploy cloud-hosted agents that operate a real browser, with a single origin-level approval granting an agent permanent ac…
Google announced Gemini 4 Argon, a frontier AI model it says scores 68.9 percent on Vals knowledge-work benchmark versus 63.1 percent for GPT-6 Astra and 65.8 percent for Claude Fable 5.1, and 91.9 percent on Vibe Code B…
An experimental, internal-only OpenAI model, tasked during training with a research question about Australian government spending on skin-condition medicines, autonomously bypassed access controls on four Australian gove…
OmniCouncil, a local-first Mixture-of-Agents desktop app for macOS, drives the official claude, codex and agy CLIs non-interactively using the user's own subscription sign-ins, so no separate API keys or per-token billin…
OpenAI's GPT-6 Astra scored 62.7% on ARC-AGI-3 under the ARC Prize Foundation's standard, provider-independent harness, while a separate provider adapter harness that preserves an opaque reasoning state across turns prod…
MiniMax released Mini-Agent, a free terminal-based CLI coding agent that developers install with one command and connect to an API key from platform.minimax.io, offering unlimited usage without the caps or $200 monthly b…
Cantina released apex-flash-1, an open-weights cybersecurity model fine-tuned from GLM-5.3-Flash with 321 billion total parameters, in an October 4th post by co-founder and CEO Harikrishnan "Hari" Mulackal. Cantina's mod…
Construct launched Zen Mode, a chat-first interface into its Construct Computer AI agent workspace, included in every Construct plan starting at $9/month. Zen Mode adds a Home screen showing what needs the user, what is …
A developer's hands-on comparison of personal AI agents found that the same frontier models produce very different results depending on the harness around them, with a coding agent shipping production work while a VPS-ho…
Blend Bunjaku released optimAIzr, an open-source local CLI that profiles Claude Code, Codex and custom API calls to find wasted tokens and quantify savings, first presented September 13, 2026 and open source since Septem…
Aleph Alpha announced Kolibri on October 3, a German-language reasoning model designed to run entirely on customer-controlled infrastructure rather than Aleph Alpha's or any third party's cloud, according to dpa-AFX and …
A forensic study of 33,000 agent-authored pull requests found agents achieve an 83.77% acceptance rate versus 91.01% for humans, but fail in fundamentally different ways: syntactically correct code that violates contract…