Opus

A new write-up from Hightouch describes an AI setup where a smaller, cheaper model handles routine work and escalates harder cases to Anthropic's Opus. According to the company, this routing approach matches Opus's accuracy while costing four times less, using LLM judges to decide when a task needs the stronger model.

A six-page introduction for graduate students to the decomposition of increasing integer sequences into weight × level + jump has been published, with the prime numbers as the main application. The method partitions sequence terms by their ratios to earlier terms, a framework used in studying prime gaps and structure. Mathematicians interested in number theory are being pointed to the report as a compact entry point to the technique.

Andre Karpathy, the AI researcher and OpenAI co-founder, says he built a working video platform called HackerTube in just two hours using Claude Opus 5.5. The feat is drawing attention as a demonstration of how quickly AI coding assistants can now take a project from idea to functioning software, and as another showcase of Anthropic's latest model capabilities.

Anthropic has updated Claude Opus 5.5 and its Claude Code tool to prevent the AI from producing unauthorized remakes of commercial video games. The change aims to stop users from using the assistant to clone copyrighted titles. The move touches on ongoing debates about AI tools and intellectual property, and is drawing attention from developers and gaming communities.

Anthropic has confirmed new usage limits on its Claude Code tool aimed at preventing users from cloning games. Developers report that the Opus 5.5 model is also refusing reverse engineering tasks, suggesting the restrictions go beyond what the company officially announced. The news is drawing criticism from some programmers who say legitimate work is being caught in the crackdown.

A developer paid 250 yuan in API fees to put China's newest large language model through its paces and compare it against Anthropic's Claude Opus. The write-up and early reactions focus on how close the Chinese model comes to Opus on coding and reasoning tasks, at a fraction of the price, feeding the wider debate over whether Chinese labs are closing the gap with leading US AI models.

Cedar Rapids pop artist Landyn Zane took the stage at Opus, according to The Gazette. The local performer played a show at the venue, drawing attention from regional coverage. Beyond the reported performance, details about the setlist, attendance or audience reaction have not been made available.

A writer reports that Claude Opus 5.5, Anthropic's latest AI model, can produce architectural drawings good enough to meet permit-level standards. The claim, based on firsthand testing described in a new write-up, has drawn attention among people following how quickly AI tools are moving into technical and professional drafting work that previously required licensed expertise.

Research firm SemiAnalysis has introduced GDPx, a new benchmark that evaluates AI agents across 11 professional domains, ranging from chip design to cybersecurity. Anthropic's Opus 5.5 tops the leaderboard, but with a score of just 12%, highlighting how far AI agents remain from expert-level performance on demanding real-world tasks. The low ceiling and breadth of tested domains are drawing attention from AI practitioners discussing how it compares with existing benchmarks and what the results say about current agent capabilities.

Anthropic has updated Claude Code so its Opus 5.5 model refuses to build direct clones of specific existing video games. The change adds instructions that stop the coding assistant from being used to recreate copyrighted titles, a move that lands amid ongoing debate about AI tools producing knockoff games and the legal grey zone around cloning gameplay and assets.

Anthropic has restricted the ability of its Claude Opus AI model to generate clones of video games, according to an IGN report. The update is said to prevent the chatbot from reproducing playable imitations of existing games, a practice that had spread widely as users prompted AI models to recreate popular titles. The move touches on ongoing debates about copyright,知识产权 and how far AI companies should police what their models produce for users.

Former Tesla and OpenAI AI researcher Andrej Karpathy says he no longer does anything resembling traditional engineering, relying instead entirely on AI coding tools such as Claude Code and Anthropic's Opus 5.5 model to build software rather than writing code himself. The remark is drawing attention as a striking example of how even experienced engineers are shifting to AI-driven development workflows.

Anthropic has rolled out an update to Claude Opus that prevents the model from generating AI-made game clones, which have flooded storefronts such as Steam. The move targets low-effort imitations of existing games produced with AI coding tools. The games industry and AI community are debating whether AI providers should police how their models are used to mass-produce derivative software.

Anthropic has rolled out an update to its Claude Opus AI model that is reported to stop it from generating AI-created game clones, the kind that have flooded platforms such as Steam. According to IGN, the change aims to curb the wave of low-effort, AI-generated games built using the model, a move likely to fuel ongoing debate over AI slop and content quality in the games industry.