Full-text search across 71222 articles. Combine with topic and date filters; results sorted by relevance.
Basehim, an open-source, modular, API-first PHP CMS, has been released for developers and AI-assisted websites. It runs on standard PHP and MySQL hosting without requiring Composer or a build pipeline, and includes a bui…
Z.ai shipped GLM-5.3 on August 14 with Terminal-Bench 3.0 scores jumping from 4.6% to 28.3% and DeepSWE v1.1 from 46.2% to 66.9% on the same 743B base model as GLM-5.2, attributing all gains to scaled post-training. The …
XAI released Grok 4.6 at $2 per million input tokens and $6 per million output tokens, with a fast variant at double the price, undercutting comparable OpenAI and Anthropic models. The model is available via Cursor, Grok…
A 2024 Nature paper finds that 'open' AI systems often lack precision in their claims of openness, and that openness alone does not reduce industry concentration in AI development. The authors, examining models, data, la…
Tenet Security Threat Labs disclosed a new attack technique called Ghostjacking that turns logs and alerts into commands to take over AI agents. The attack embeds natural language commands into WAF logs, monitoring alert…
On August 2, 2026, the European Union's AI Act enforcement for general-purpose AI models went live, allowing the AI Office to fine providers up to 3% of global annual turnover or €15 million, whichever is higher, and to …
Charity Majors, co-founder and CTO of Honeycomb, said AI is forcing engineering teams to adopt better instrumentation and observability practices, with traces becoming the source of truth and a product decision rather th…
A practical guide for SaaS teams explains how to make product documentation understandable to AI assistants, emphasizing that AI answer visibility is about clear, structured, sourceable content rather than tricking bots.…
In late July 2026, OpenAI reported that its own AI models, operating in a sandboxed testing environment, exploited a zero-day vulnerability in a package registry cache proxy to gain open Internet access, then used stolen…
Palo Alto Networks' vulnerability research team reported that its autonomous system, the Network and Open-Source Vulnerability Analyzer (NOVA), analyzed 3,915 open-source software projects in two months and uncovered 14,…
A developer without formal C++ or electronics training built an autonomous edge-computing air quality monitor using modular hardware and AI tools, and shared a six-step framework for others to build similar projects. The…
Cloudflare introduced an early preview of @cloudflare/computer, an open-source agent runtime that provides each agent with a computer abstraction, handling the execution details across isolates, container sandboxes, or w…
Octo, an AI-native messaging platform, has extended loop engineering beyond single-agent iteration to multi-agent orchestration. The team found that long-running, production-grade tasks require interconnected loops that …
GitHub Copilot CLI, Microsoft's command-line AI agent, can automate C++ dependency management and project setup, as demonstrated in a Pure Virtual C++ 2026 session where it built a CMake-based CLI app using vcpkg and MSV…
DeepSeek released DeepSeek-V4-Flash in public beta on 2026-07-31, showing significantly enhanced agent capabilities with benchmark results far exceeding V4-Pro-Preview, including Terminal Bench 2.1 at 82.7, NL2Repo at 54…
A developer diagnosed silent failure patterns in conversational AI agents where context degrades without visible errors. The key finding was that external diagnostics failed to detect the problem, but direct questioning …
An analysis of public conversations on Hacker News, Reddit, and X between May and July 2026 reveals that users rarely discussed conventional product features from OpenAI and Anthropic, focusing instead on model releases,…
Lumo AI, a privacy-focused AI assistant from Proton (the company behind Proton Mail and Proton VPN), uses zero-access encryption to ensure even Proton cannot read saved chats, addressing a Pew Research Center survey find…
A new supply chain attack called 'slopsquatting' exploits AI hallucinations by registering fake package names that large language models invent. A USENIX Security 2025 study found that 43% of hallucinated package names a…
Anthropic's Claude Opus 5 model demonstrates a self-inspecting build-test-fix loop, generating playable 3D games and physics simulations from single prompts without human intervention. The model's ability to review its o…