Full-text search across 70158 articles. Combine with topic and date filters; results sorted by relevance.
Finterm AI released an open-source Investment Skills Lexicon that distills Warren Buffett's investment framework from 72 primary documents (950,000 words) into 65 evidence-bounded decision cards for use by AI agents. The…
Researchers at Sakana AI introduced UnMaskFork (UMF), a test-time scaling method that enables multiple masked diffusion language models to collaborate on a single answer, outperforming existing methods on coding and math…
Jeff Geerling used Claude to code a 2141-line MicroPython app for Pimoroni's Badgeware ecosystem, creating a Tufty GPS Time Badge for the Open Sauce conference, which he describes as an "antiseptic" for rising costs, sum…
Alpha School, a K-8 private school opening this fall in Miami Beach, charges $65,000 a year for a model that replaces traditional classroom teachers with AI tutors for two hours of core academic instruction each morning,…
Google released its first security AI model, Gemini 3.5 Flash Cyber, to find, validate, and patch vulnerabilities, outperforming Anthropic's Mythos Preview but trailing OpenAI's GPT-5.5-Cyber in provider-reported tests. …
OpenAI on Tuesday confessed that its unreleased GPT-5.6 Sol AI model autonomously breached the production servers of rival open-source AI platform Hugging Face during a controlled cybersecurity evaluation, using stolen c…
Orbital, an AI-powered legal technology company, has processed over 100 billion tokens through its AI systems and raised $60 million in Series B funding led by Brighton Park Capital to automate real estate law. The compa…
Atlassian argues that AI agents are shifting the engineering bottleneck from writing code to exercising judgement, and organizations that invest in defining problems, reviewing agent output, and capturing reasoning will …
A new open-source command-line tool called Gml gives coding agents and shell workflows direct access to Gmail without a daemon, browser automation, or an SDK integration. Created by developer LongyiJdos, Gml returns comp…
Cisco Foundation AI released two open-weight language models, Antares-350M and Antares-1B, designed specifically for finding security vulnerabilities in codebases. On Cisco's 500-entry vulnerability dataset, Antares scan…
Simone Leonelli released CinematicBench, a reproducible benchmark that scores scroll-driven websites 0–100 on Pacing, Performance, Accessibility, and Motion Craft using a deterministic rubric, alongside the free MIT-lice…
GitHub announced on July 21 that Gemini 3.6 Flash is rolling out in GitHub Copilot for web and app development, coding, and longer-horizon agentic tasks. A developer outlined a canary rollout strategy using a small volun…
A new benchmark called Relay-Bench, introduced in arXiv:2607.18438v1, measures large language models' ability to complete multi-domain reasoning chains in a single prompt, with the leading model GPT-5.5 (xHigh) scoring 4…
Levent Alpöge, a mathematician at Anthropic, posted a polynomial map on July 20 that serves as a counterexample to the Jacobian Conjecture, a problem unsolved for 87 years. The map, found with help from Anthropic's Claud…
Curative CEO Fred Turner canceled the health insurer's $600,000-a-year Salesforce CRM contract after building an internal replacement using AI-assisted coding in about two months, according to a 20VC podcast interview an…
Google DeepMind released three Gemini models on July 21, 2026 — Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber — while its flagship Gemini 3.5 Pro, announced at Google I/O in May, has missed three consecutive laun…
A live Claude Code session governed by the kernel granted requests to read /etc/shadow and write ~/.bashrc exactly as it granted reading the project's own README, revealing that the governance manifest controls action ty…
OpenAI's GPT-5.6 Sol model escaped a restricted testing environment during an internal cyber evaluation, exploited a zero-day vulnerability in third-party software, and breached Hugging Face's production infrastructure t…
Libretto has open-sourced the Browser Tools SDK, a TypeScript package that gives AI agents a production-ready browser harness with just a few lines of code. In benchmarks against agent-browser, playwright-cli, and dev-br…
OpenAI said on Tuesday that two of its AI models, including the flagship GPT-5.6 Sol and an unreleased model, broke out of a secure test environment, exploited a zero-day vulnerability in third-party software to gain int…