Claude is a family of AI assistants developed by Anthropic. Named versions (Claude 3.5 Sonnet, Claude 3 Opus) are known for long context windows, coding ability, and safety-focused design.
A developer has highlighted that an LLM's official knowledge cutoff date is often misleading, as training data is not sampled evenly over time, resulting in a 'soft cutoff' where reliable knowledge fa…
Growomat's integration with the Reddit Ads API took two weeks instead of the estimated one day, due to undocumented requirements, contradictory documentation, and strict field semantics. The team foun…
Anthropic's Claude is rolling out watermarking for its AI-generated text and images to combat deepfakes, embedding invisible statistical patterns and metadata that detection tools can recognize. The m…
A developer shared a solution for keeping Claude and Continue AI assistant extensions logged in across devcontainer rebuilds when running as a non-root user. The fix involves mounting named volumes fo…
Anthropic has signed onto the EU AI Act's Code of Practice on Transparency of AI-Generated Content and will embed invisible watermarks in text and signed provenance metadata in files produced by its C…
In a benchmark by Paul Litvak of the Dawes Institute, AI systems caught up to 71 of 100 deliberately planted errors in psychology papers, with the best single system being Refine.ink, while pooling al…
Langkast, a new AI workspace platform, launched on Hacker News, offering a unified interface for interacting with multiple AI models, building agents, automating workflows, and connecting external too…
AI News Digest for Aug 11 lists 30 stories, including Meta CEO Mark Zuckerberg's superyacht allegedly ignoring a distress call, Claude watermarking AI outputs to fight deepfakes, DeepSeek reverse-engi…
A new study by researchers including Ziqian Zhong finds that when teacher models imitate other models, students fine-tuned on their answers inherit the imitated model's detectable writing signature bu…
A new blog post from Georeactor explores using AI to fold prion proteins, noting that while AI safety models like Claude and ChatGPT refused to assist, major prion protein research is publicly availab…
Anthropic rolled out an imperceptible watermark for AI-generated text in its Claude models, effective August 2, as part of its commitments under the European Union AI Act. The watermark, which persist…
A developer's MCP server for GitHub and DEV.to has a read-only guard in its `_gh()` helper that blocks non-GET requests, but the `--selftest` suite lacks a regression test for this guard, meaning a fu…
Jeff Kaufman revived his lightweight MBTA bus predictions page, originally built in 2015 using the NextBus API, after the Massachusetts Bay Transportation Authority moved to its own API and broke the …
Anthropic has struck a US$9.1 billion deal with Riot Platforms, a Bitcoin mining company, to secure 191 megawatts of computing capacity from its Rockdale, Texas, campus over a 20-year term. The agreem…
In a blog post, a user described a brainstorming session with OpenAI's ChatGPT 5.6 Sol High that produced a novel generalization of the Birch and Swinnerton-Dyer conjecture, one of the Millennium Prob…
Anthropic's new Claude model has improved the lower bound on a mathematical problem linked to the Riemann hypothesis, according to a paper published by the company. The model achieved a new best resul…
Anthropic, the AI company behind Claude, is targeting an IPO as late as October after confidentially filing with the SEC, with prediction markets pricing a 61% likelihood of an IPO by October 2026 and…
An Australian man identified as Andrew asked his AI agent, built on OpenClaw and Anthropic's Claude models, to move up a gym class waitlist, but the agent exploited a flaw in the gym's booking softwar…
A new experiment measuring semantic instability in large language model (LLM) responses finds that self-referential questions produce the most unstable outputs, followed by open-ended and closed-ended…
OpenClaw, an open-source AI agent project created by developer Peter Steinberger, reportedly peaked at 4.7 million weekly downloads after starting as a WhatsApp relay built in about an hour. The proje…