Full-text search across 24325 articles. Combine with topic and date filters; results sorted by relevance.
President Trump abruptly canceled a planned executive order that would have established voluntary pre-deployment safety evaluations for frontier AI models, hours before the scheduled signing ceremony. The decision follow…
Forge is a reliability layer for self-hosted LLM tool-calling that uses guardrails and context management to dramatically improve performance on multi-step agentic tasks, boosting an 8B model from 53% to 99% accuracy. It…
Anthropic's "Dreams" feature is not merely a personalization tool but an asynchronous memory consolidation pipeline that processes user data during low-demand periods to improve GPU utilization and reduce costs. It descr…
Polytoken 0.8.20 added the claude-haiku-5.5 model to its Anthropic provider, cleaned up rate-limiting logic, and made the rate-limit timer honor a provider's Retry-After header. The release also lets the ask_user_questio…
A group led by Geoffrey Hinton, Stuart Russell and Arvind Narayanan published a public letter setting five minimum standards for credible embedded evaluators at frontier AI companies, including independence from continge…
OpenAI, Anthropic and Google plan to launch a "Standards Authority for Frontier AI" that would set incident reporting standards and establish qualifications for independent auditors of models and labs, according to The I…
Venture capitalist David Sacks has become the leading voice advising President Donald Trump against government AI safety regulation, with a senior White House official telling Politico that Trump's newly vocal hostility …
RuntimeWire found that Kimi Work 3.2.7 copied a complete user prompt verbatim into a kimi_work_send_message product-analytics event sent to gator.volces.com, a Volcano Engine DataFinder reporting endpoint. The finding ca…
A Python toolkit that pushes a data question through six stages—business understanding, hypothesis generation, SQL planning, validation, executive summary, and recommendations—aims to make AI data analysts check their nu…
Anthropic removed over 80% of Claude Code's system prompt for the Claude 5 generation with no measurable loss on coding evaluations, while an ICLR 2026 paper found that context collapse from rewriting context causes shar…
Alibaba's Qwen team released Qwen3.8-Flash-Next, an open-weight multimodal mixture-of-experts model previewing the architecture for the upcoming Qwen4. The model features 125 billion total parameters with only 6 billion …
Anthropic introduced a watermark into Claude's output this month, and others have followed or will soon follow suit. The watermark, similar to Google's SynthID, uses statistical bias from a pseudorandom generator when sa…
Anthropic's Thariq Shihipar confirmed that Claude Pro and Max usage limits are token-based, not message-based, and that the 5-hour session window can drain faster during peak hours, affecting about 7% of users. The compa…
Anthropic's Claude Code can now run on-premises with AMD Instinct GPUs, serving GLM 5.2 at full quality via SGLang and LiteLLM, eliminating cloud API dependency and per-token costs. The setup uses an SSH tunnel to connec…
Missing Studio released Eva 0.1.0, an open-source, multi-tenant, AI-native software factory that logs every interaction to disk as evidence, with the terminal chat client showing only what is already recorded. The tool, …
Leopold Aschenbrenner, a 24-year-old former OpenAI researcher and founder of hedge fund Situational Awareness LP, was forced to sell all public stock holdings after the fund posted large losses from declines in AI infras…
Dynamic tool discovery, which lets AI agents search for the right tool instead of loading all schemas, does not solve the core problems of cost, latency, and governance because retrieval often fails to find the correct t…
Open weights AI models, predominantly from China, threaten the dominance of frontier labs like OpenAI and Anthropic, which have grown revenues rapidly but may be disrupted before capitalizing on their technology. The ris…
AMD CEO Lisa Su said the total addressable market for AI products will grow 45% a year through 2030 to $1.4 trillion, up from her previous estimate of $500 billion six months ago, driven by surging demand for compute fro…
Anthropic released Claude Opus 5 on July 24, calling it a step-change improvement over Opus 4.8. The model approaches Fable 5 intelligence at roughly half the price, with a 1M token context window and 128k max output tok…