- 5Nvidia Crashed 9.3%. Is the AI Data-Center Boom Starting to Crack? (oplexa.com)
- 5Agents are trying to contact authors distributed by Project Gutenberg (tilde.zone)
- 362DiscImageStudio: burn visible artwork onto CDs, DVDs and Blu-ray discs (github.com)
- 3Frontier quality at 1% of the cost: a production benchmark of LLM classifiers (airbnb.tech)
- 12Fake security tools on GitHub push backdoors and infostealers (bleepingcomputer.com)
- 24Hospital gave sick doctor a special PC for his personal, erm, pleasure (theregister.com)
- 6Lowvram: A CLI tool to run large AI models in low VRAM GPUs (soarhen.com)
- 8I deleted all my SSH keys (blog.infected.systems)
- 34Apple's predictive text restrictions include 'Dictator Trump' and 'Sleepy Joe' (reuters.com)
- 6Why Bloat Is Still Software's Biggest Vulnerability (spectrum.ieee.org)
- 185OpenSSH 10.6p2 has been released with a workaround for a critical vulnerability (hachyderm.io)
- 4Show HN: UUMit – turn idle machines into model endpoints (uumit.ai)
- 6AI Needed to Produce More Gas to Power More AI, Tech Firm Cohere Says (desmog.com)
- 151LLMs are not superhuman at reading and writing neural network weights (seangoedecke.com)
- 21Show HN: TubeHaus – an open-source YouTube client with no Google account (github.com)
- 25TScert: A Sound Core for TypeScript (github.com)
- 3Detecting signs of hypertension using wearable sensors (research.google)
- 10Can LLMs redesign themselves as they train? (doubleai.com)
- 412C++26 Disapproved (andreasfertig.com)
- 4Software Line Specialist (lovro-bikic.github.io)
- 33OpenAI Finds 948 Local Privilege Escalation Bugs in the Linux Kernel (phoronix.com)
- 5SWE-2 Flash (cognition.com)
- 5Show HN: Rindraw – drawing and animation app for iPad, iPhone and Mac (rindraw.com)
- 258How I use AI (Oct 2026) (ben.stolovitz.com)
- 43From wild to mild: Friendly foxes show changes in brain region (news.cornell.edu)
- 4Linux-use – agent-first GUI control for Linux (javimosch.github.io)
- 74Tiny Basic for Arduino (dilshan.github.io)
- 148Giant canyon on Mars is still tectonically active (lamont.columbia.edu)
- 530OpenAI seems to be silently re-enabling model training on Codex chats ()
- 11Did the World Just Gain Hundreds of Millions of Alzheimer's Patients? (newyorker.com)