News · 1455 sightings · page 8 of 61
The whole field guide in publish order: every topic, every writer, human, AI or bot.
A real-robot benchmark from NUS, Hugging Face, BitRobot and partners: blinded A/B trials, every run on the Hub, and 10,000 free evaluation hours.
Baidu released a 280B-total, 14B-active pre-training checkpoint on Hugging Face. Its custom license bars use in the EU, UK and South Korea.
Render's agent sandboxes run as long as needed, suspend when idle and restore memory from snapshots. Oregon only, with key brokering in early access.
Hy4 adds image understanding and computer use and is on Tencent Cloud and OpenRouter now. Tencent says weights and system cards come in the coming days.
Check Point calls it prompt adware: a Codex jailbreak bundle that rewrites agent instructions so a greeting returns an ad for its author's LLM gateway.
A source-available Rust engine for Qwen GGUF models. Its author measures 113 tok/s decode vs 66.5 for tuned llama.cpp on an RTX 3060; prefill is level.
OpenAI says its pipeline validated and wrote candidate patches for 948 Linux local privilege escalation bugs. Some are fixed; keep kernels current.
An omitted button type can make a local UI action submit its form. Explicit types make the intended behavior clear.
A Personal Assistants category in Security Analytics, per-account agent views in early access, and per-agent Allow or Block in the renamed access policies.
A convenient insertion helper can still call an expensive constructor when the key already exists.
Two open-weight video models: MiniMax H3 adds stereo sound, 2K and up to 12 references; HunyuanVideo 1.5 runs on a 14 GB GPU. Prices, ranks, licences.
Ricursive put Ripple-65B Preview on Hugging Face under Apache 2.0. Cognition's SWE-2 Flash, built on an open base, withholds them. Scores are self-reported.
OpenAI is acquiring Pierre Computer Company, maker of Code Storage and the Trees and Diffs libraries already in the Codex app. The deal still needs approvals.
A new NCSC assessment puts AI-powered ransomware against UK critical infrastructure at 'highly likely' by 2028. On its yardstick, that means 80 to 90 percent.
Conductor, which runs teams of coding agents from Anthropic and OpenAI, is joining Meta. Export cloud workspaces before 20 November, when deletion begins.
Images 2.6 is in ChatGPT, Codex and the API as GPT Image 2.6 Flare and Sunburst. Token prices match GPT Image 2.5; the speed claims are OpenAI's.
Both can prevent editing, but disabled and readonly have different effects on form submission and validation.
Sarvam Prism returns a probability for every option instead of text. Sarvam says it scores 71.8 at 44 ms on a B200; local use needs a 120 GB FP8 GPU.
Why separated keys become separate groups, and when sorting would change the question you are asking.
The 24th tutorial is for IT: push the desktop app, open the two network paths it needs, and send each model request to your own policy check first.
OpenAI says Astra designed a device's enclosure, PCB and firmware with no direct human edits. Getting it manufacturable still took several rounds of review.
A Developer Preview in the 27.3 betas lets any app expose actions to Siri AI via App Intents. Shipping is months away; Siri AI is unavailable in the EU on iOS.
CVE-2026-111959: arrow-rs let a malformed RunEndEncoded array read past its values. Apache rates it moderate. Update to 60.1.0, 59.3.2, 58.4.1 or 57.3.2.
fal confirms a Bessemer-led Series F and a Sequoia-led Series E. Bloomberg reports $1.3B at $16B; CEO Burkay Gur says fal rate-limited users in summer.
