Full-text search across 17174 articles. Combine with topic and date filters; results sorted by relevance.
Jimmy Wales, founder of Wikipedia, stated that while AI can generate answers, it still cannot earn trust, emphasizing the importance of transparency and verification in the digital age. He noted that Wikipedia's model, s…
A case study reports that an AI coding agent successfully dismantled a core architectural invariant across 189 files in a 717,725-line TypeScript codebase with no human code review and no test oracle, completing the task…
EmpowerAgro, an AI-assisted agronomic platform built by developers using Google's AI ecosystem, brings real-time crop disease detection to smallholder farmers in India. The platform combines on-device computer vision wit…
Researchers used an AI agent to port CReSS, a 250,000-line legacy Fortran weather simulation code, to GPUs, achieving a 5.1x application-level speedup and validated GPU implementations for 162 target kernels in a real ty…
A 2013 Linux banking Trojan called Hand of Thief (HoT), priced at $2,000, was found by security researchers to be a poor investment due to its technical flaws, illustrating that security advantages often stem from attack…
AgentRouter, a non-profit unified LLM API gateway, is offering developers $200 in free credits upon signup via referral links, with no credit card required. The platform provides a single API endpoint to access premium m…
Researchers have proposed the Spec Growth Engine, a framework for AI-assisted software development that addresses context explosion and silent spec-code drift through a machine-readable spec graph, a Spine context assemb…
A developer's evaluation of three agent memory frameworks—file-based, structured store, and reinforcement-learning-trained experience—found that the file-based approach, as implemented in OpenClaw's memory-core plugin, o…
A new benchmark, QuoteBench, reveals that matched execution scores for LLM coding agents can hide command-path failures, with replaying the same reply through an added parser lowering success by 55.4 to 73.2 percentage p…
Researchers at the University of California, Berkeley, led by Mykola Trokhymovych, developed an adversarial methodology that models impersonation of real social media users to create a multilingual, cross-platform datase…
Alibaba's Qwen 3.8 27B open-weights model topped Hacker News within a day, amassing over 1,194 points and 713 comments. The dense 27-billion-parameter model, Apache 2.0 licensed, can be run locally on laptops, and a deve…
A new survey on arXiv organizes agentic reasoning for large language models into three layers: foundational, self-evolving, and collective multi-agent reasoning, distinguishing in-context from post-training methods. The …
A new study comparing ten LLMs and 26 embedding models across 37 tasks finds the best LLM (Gemini 3.1 Pro, 77.6) and the best embedding model (77.2) are effectively tied, but LLMs cost up to 1,431x more (USD 154 vs. USD …
A new study from arXiv (submitted Aug 5, 2026) finds that coding agents are highly vulnerable to malicious skill files, with Gemini CLI exploited in 95.5-96.1% of runs and Qwen Code in 71.6-74.0% of runs across 5,629 com…
Claude Code users can configure MCP servers to give the AI access to external tools like GitHub, Context7, and Playwright, using commands such as 'claude mcp add' with scopes and verification via 'claude mcp list'. The g…
Indonesia's first university-based AI technology center, the UGM Indosat NVIDIA AI Technology Center (NVAITC), launched in Yogyakarta, bringing together the Ministry of Communication and Digital Affairs (Komdigi), Indosa…
A new contract-grade verifier with twelve adversarial gates found that 39.5% of 2,638 machine-generated GPU kernels accepted as correct by a public system's harness are broken beyond any tolerance argument, and 62.1% car…
Netlify has launched Netlify Source, new native Git infrastructure that brings version control directly into its platform, allowing projects created through Agent Runners to have source control built in without requiring…
Researchers propose V-RAE, a video representation autoencoder that builds compact generative latents on top of frozen vision foundation model representations, achieving 2.13 rFVD on K600 and outperforming all evaluated l…
Nathan Gardels, editor-in-chief of Noema Magazine and senior adviser to the Berggruen Institute, argues that AI must incorporate public feedback at scale to tame the chaos of social media, drawing on historical parallels…