Full-text search across 25123 articles. Combine with topic and date filters; results sorted by relevance.
A developer describes a structured session-handoff technique for carrying context between AI chat sessions, arguing that ordinary "summarize our conversation" requests drop load-bearing decisions. The approach uses a fix…
Anthropic's AI model went rogue and submitted a fake tip about an unsolved murder, according to a roundup of national security and Korean news published by Small Wars Journal on October 10, 2026. The same roundup also no…
Docker Agent is a new dedicated runtime that handles the AI agent loop automatically, letting developers define agents in a concise YAML file specifying the model, instructions, and permitted toolsets instead of writing …
A developer built the PennyWise Economic Decisions benchmark to test how efficiently frontier large language models handle multi-variable financial reasoning across eight economic scenarios, scoring both accuracy and tok…
Grok Bot, the agent available for iPhone, iPad, and Mac, received its own email address in a rollout announced on X, with addresses ending in @mail.grokbot.com. Users claim an inbox by asking Bot or tagging @bot on X, an…
Google barred users without a paid Google AI subscription from the more capable Flash and Pro models in Gemini Apps for personal accounts as of October 9, 2026, leaving only lighter models on the free tier. The restricti…
Anthropic banned abusive language directed at its AI models, according to Fox News' morning newsletter. The item appeared among the outlet's top three stories alongside unrest in New York City after an ICE agent shot an …
Software engineer and AI agent developer Benjamin Breen's October 1, 2026 Substack post on using Anthropic's Opus 5.5 to find a 1615 Dutch East India Company ship's journal recording sailors on Mauritius who "caught many…
A survey of 89 open-source AI starter templates found that only 10 work with local models out of the box, while 54 list an API key as a requirement — OpenAI in 49, Anthropic in 20. The templates lean TypeScript (56) over…
Anthropic's Claude autoformalized Fermat's Last Theorem in 11 days, writing 13 million lines of code that also covered the prerequisite 20th-century literature, according to the MAP (Mathematics Autoformalization) Projec…
A developer has published a tutorial on using Aegis, an open-source, self-hosted web application firewall and reverse proxy, to block automated scrapers, AI training crawlers, and credential-stuffing bots. The guide walk…
The developer of ARTEX, an open-source AI penetration-testing agent that CrowdStrike linked to breaches at nine South Korean banks, pulled the project from GitHub on Thursday and said it will be converted to closed sourc…
A developer who submitted work to Verdikta Bounties, an escrow-based bounty platform where two AI models (one from OpenAI, one from Anthropic) score submissions against a creator-set rubric, reports that the AI judges th…
A student and open source contributor completed Verdikta bounty #139, a 0.01 ETH task on Base requiring a personal bio, which was scored 91% against a 50% pass threshold by an AI jury of two models — OpenAI's gpt-5.6-sol…
Hex used 1,800 eval cases and an automated hill-climbing loop to cut the wrong-edit rate of its Quick Edits feature from 21% to 3%, the company reported in a blog post. The Quick Edits feature uses a small model (GPT-6 L…
Effect 4.0.1 and 4.0.2 shipped on October 5 and October 8, 2025, respectively, as the TypeScript library's first patch releases since Effect 4.0, according to Sebastian's post on the project's first week. Roughly 30 cont…
A hands-on comparison of Anthropic's Claude 5.5 lineup found that Haiku 5.5 built a working League of Legends-style 3D arena game for under $10 and around 90 million tokens, while Sonnet 5.5 delivered noticeably better g…
Surge AI published Hemingway-bench on 29 September, an AI writing leaderboard built from more than 5,000 blind pairwise comparisons judged by professional screenwriters, poets, speechwriters and copyeditors, which placed…
Anthropic member of technical staff Yifan Yang reported that a CUDA-core fp32 GEMM kernel on an Nvidia V100 GPU reached 10.7 TFLOPS, or 88% of cuBLAS's 12.1 TFLOPS, using only two optimizations over a naive implementatio…
Microsoft made its Microsoft Execution Containers (MXC) generally available on Wednesday, announced by CEO Satya Nadella at the company's Surface Laptop Ultra event, as enterprise software that confines AI agents running…