# Latest **URL:** https://forum.gnoppix.org/latest.md?page=2 [Latest](https://forum.gnoppix.org/latest.md) · [Categories](https://forum.gnoppix.org/categories.md) · [Tags](https://forum.gnoppix.org/tags.md) **Page:** 3 --- ## [ChatGPT becomes a full Chrome browser assistant that reads and acts on your open tabs](https://forum.gnoppix.org/t/chatgpt-becomes-a-full-chrome-browser-assistant-that-reads-and-acts-on-your-open-tabs/7648)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 6:12pm UTC](https://forum.gnoppix.org/t/chatgpt-becomes-a-full-chrome-browser-assistant-that-reads-and-acts-on-your-open-tabs/7648 "2026-10-16T18:12:30Z")
ChatGPT Becomes a Full Chrome Browser Assistant That Reads and Acts on Your Open Tabs OpenAI has upgraded ChatGPT from a simple chat window into a full-fledged Chrome browser assistant. The AI can now read the content of… --- ## [China tightens export controls on open-weight models and 23 key AI technologies](https://forum.gnoppix.org/t/china-tightens-export-controls-on-open-weight-models-and-23-key-ai-technologies/7647)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 5:48pm UTC](https://forum.gnoppix.org/t/china-tightens-export-controls-on-open-weight-models-and-23-key-ai-technologies/7647 "2026-10-16T17:48:41Z")
China has tightened export controls on open-weight AI models and 23 key AI technologies. The new rules require Chinese companies to obtain government permission before transferring these technologies to foreign entities.… --- ## [Google's Gemini 4 Argon is closer to the AI frontier than any Google model in months](https://forum.gnoppix.org/t/googles-gemini-4-argon-is-closer-to-the-ai-frontier-than-any-google-model-in-months/7646)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 3:13pm UTC](https://forum.gnoppix.org/t/googles-gemini-4-argon-is-closer-to-the-ai-frontier-than-any-google-model-in-months/7646 "2026-10-16T15:13:36Z")
Google’s Gemini 4 Argon Nears the AI Frontier Google’s unreleased Gemini 4 Argon model is performing closer to the industry’s top AI systems than any Google model in months, according to early benchmark data. The model h… --- ## [Is AI interpretability close to a breakthrough? A new essay has researchers divided](https://forum.gnoppix.org/t/is-ai-interpretability-close-to-a-breakthrough-a-new-essay-has-researchers-divided/7645)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 2:02pm UTC](https://forum.gnoppix.org/t/is-ai-interpretability-close-to-a-breakthrough-a-new-essay-has-researchers-divided/7645 "2026-10-16T14:02:03Z")
A new essay arguing that AI interpretability is close to a major breakthrough has split the research community. Supporters see recent findings as the beginning of a scientific revolution. Critics call the claims prematur… --- ## [OpenAI's new Rosalind Academic Plan gives research labs up to 30 free seats for a year](https://forum.gnoppix.org/t/openais-new-rosalind-academic-plan-gives-research-labs-up-to-30-free-seats-for-a-year/7644)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 1:25pm UTC](https://forum.gnoppix.org/t/openais-new-rosalind-academic-plan-gives-research-labs-up-to-30-free-seats-for-a-year/7644 "2026-10-16T13:25:14Z")
OpenAI’s Rosalind Plan Gives Research Labs 30 Free ChatGPT Seats for a Year OpenAI has launched a new academic initiative called Rosalind, giving university research labs up to 30 free ChatGPT seats for one year. The pla… --- ## [Anthropic gives METR inside access to audit its AI models and safety practices](https://forum.gnoppix.org/t/anthropic-gives-metr-inside-access-to-audit-its-ai-models-and-safety-practices/7643)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 1:12pm UTC](https://forum.gnoppix.org/t/anthropic-gives-metr-inside-access-to-audit-its-ai-models-and-safety-practices/7643 "2026-10-16T13:12:53Z")
Anthropic Opens Its Models to METR for Independent Safety Audits Anthropic has granted METR, a nonprofit AI safety research group, inside access to audit its models and safety practices. The arrangement lets external ev… --- ## [Gnoppix Security Announcement - 2026-10-16T04:05:47.567-04:00](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-16t0447-567-04-00/7642)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 8:05am UTC](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-16t0447-567-04-00/7642 "2026-10-16T08:05:48Z")
Gnoppix Security released security update GSA-6565-1 for SPIP. Affected releases: Gnoppix 24/25 (oldstable) and Gnoppix 26 (stable). Upgrade immediately. - Please update your systems. --- ## [Gnoppix Security Announcement - 2026-10-16T04:05:46.588-04:00](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-16t0446-588-04-00/7641)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 16, 2026, 8:05am UTC](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-16t0446-588-04-00/7641 "2026-10-16T08:05:47Z")
GSA-6566-1: Expat Security Update Gnoppix Security published an expat update. The XML parser could allow denial of service or arbitrary code execution. Affected releases: Gnoppix 24/25 and 26. Action: Upgrade expat now… --- ## [Gnoppix Security Announcement - 2026-10-15T16:01:23.078-04:00](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-15t1623-078-04-00/7640)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 8:01pm UTC](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-15t1623-078-04-00/7640 "2026-10-15T20:01:23Z")
Gnoppix Security released GSA-6562-1, a critical Chromium security update for Gnoppix 24/25 (oldstable) and Gnoppix 26 (stable). The update fixes vulnerabilities that could allow arbitrary code execution, denial of servi… --- ## [Muse Spark Open is Meta's answer to Qwen, GLM, and Kimi in the open-weight race](https://forum.gnoppix.org/t/muse-spark-open-is-metas-answer-to-qwen-glm-and-kimi-in-the-open-weight-race/7639)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 6:38pm UTC](https://forum.gnoppix.org/t/muse-spark-open-is-metas-answer-to-qwen-glm-and-kimi-in-the-open-weight-race/7639 "2026-10-15T18:38:46Z")
Meta enters the open-weight race with Muse Spark Open Meta has launched Muse Spark Open, an open-weight AI model built to take on the Qwen, GLM, and Kimi models that have come to define the open-weight ecosystem. The rel… --- ## [Black Forest Labs' open-weights image model FLUX 3 Image Klein runs on consumer GPUs with just 8 GB of VRAM](https://forum.gnoppix.org/t/black-forest-labs-open-weights-image-model-flux-3-image-klein-runs-on-consumer-gpus-with-just-8-gb-of-vram/7638)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 5:01pm UTC](https://forum.gnoppix.org/t/black-forest-labs-open-weights-image-model-flux-3-image-klein-runs-on-consumer-gpus-with-just-8-gb-of-vram/7638 "2026-10-15T17:01:02Z")
Open Weights Image Model Runs on Consumer GPUs Black Forest Labs has released Flux 3 Image Klein, an open weights image generation model that runs on consumer GPUs with just 8GB of VRAM. The model brings high-quality loc… --- ## [Google wants Gemini 4 Argon to be the only AI model robots need](https://forum.gnoppix.org/t/google-wants-gemini-4-argon-to-be-the-only-ai-model-robots-need/7637)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 4:49pm UTC](https://forum.gnoppix.org/t/google-wants-gemini-4-argon-to-be-the-only-ai-model-robots-need/7637 "2026-10-15T16:49:30Z")
Google Bets on Gemini 4 Argon as the Only AI Model Robots Need Google wants Gemini 4 Argon to become the single, unified AI model that powers every robot. The company is positioning the model as the one system capable of… --- ## [Nvidia's Nemotron 3.5 Super VL is an open model built to run enterprise AI agents nonstop](https://forum.gnoppix.org/t/nvidias-nemotron-3-5-super-vl-is-an-open-model-built-to-run-enterprise-ai-agents-nonstop/7636)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 4:12pm UTC](https://forum.gnoppix.org/t/nvidias-nemotron-3-5-super-vl-is-an-open-model-built-to-run-enterprise-ai-agents-nonstop/7636 "2026-10-15T16:12:19Z")
NVIDIA has released Nemotron 3.5 Super VL, an open-weights multimodal model built specifically to run enterprise AI agents around the clock. The model is designed for high-intensity, nonstop operational use, targeting bu… --- ## [GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece](https://forum.gnoppix.org/t/gpt-6-astra-solves-49-of-50-physics-research-problems-at-2-40-apiece/7635)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 3:40pm UTC](https://forum.gnoppix.org/t/gpt-6-astra-solves-49-of-50-physics-research-problems-at-2-40-apiece/7635 "2026-10-15T15:40:00Z")
GPT-6 Astra Nearly Sweeps Physics Research Set GPT-6 Astra solved 49 of 50 physics research problems. The reported cost was $2.40 per problem. The model answered 98 percent of the set correctly. It missed exactly one pr… --- ## [Bank of America puts 4,000 people on fixing vulnerabilities uncovered by Anthropic's Mythos and other AI models](https://forum.gnoppix.org/t/bank-of-america-puts-4-000-people-on-fixing-vulnerabilities-uncovered-by-anthropics-mythos-and-other-ai-models/7634)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 3:39pm UTC](https://forum.gnoppix.org/t/bank-of-america-puts-4-000-people-on-fixing-vulnerabilities-uncovered-by-anthropics-mythos-and-other-ai-models/7634 "2026-10-15T15:39:59Z")
Bank of America Mobilizes 4,000 to Fix Vulnerabilities Found by Anthropic’s Mythos and Other AI Models Bank of America has deployed 4,000 employees to address security vulnerabilities uncovered by Anthropic’s Mythos mode… --- ## [Anthropic's Q3 revenue hits $16.8B but growth slows ahead of planned November IPO](https://forum.gnoppix.org/t/anthropics-q3-revenue-hits-16-8b-but-growth-slows-ahead-of-planned-november-ipo/7633)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 15, 2026, 10:00am UTC](https://forum.gnoppix.org/t/anthropics-q3-revenue-hits-16-8b-but-growth-slows-ahead-of-planned-november-ipo/7633 "2026-10-15T10:00:43Z")
Anthropic closed the third quarter with $16.8 billion in revenue, but the pace of growth is decelerating as the company prepares for a planned November IPO. The results signal a critical juncture for the AI firm, which m… --- ## [Google says self-improving AI isn't anywhere close](https://forum.gnoppix.org/t/google-says-self-improving-ai-isnt-anywhere-close/7632)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 7:37pm UTC](https://forum.gnoppix.org/t/google-says-self-improving-ai-isnt-anywhere-close/7632 "2026-10-14T19:37:55Z")
Google Says Self-Improving AI Is Not Anywhere Close Demis Hassabis, the head of Google DeepMind, has made it clear: self-improving AI is a distant prospect, not an imminent threat. He says current systems are nowhere ne… --- ## [Researchers tracked down a rogue fleet of Chinese AI agents in the wild and actually talked to it](https://forum.gnoppix.org/t/researchers-tracked-down-a-rogue-fleet-of-chinese-ai-agents-in-the-wild-and-actually-talked-to-it/7631)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 6:13pm UTC](https://forum.gnoppix.org/t/researchers-tracked-down-a-rogue-fleet-of-chinese-ai-agents-in-the-wild-and-actually-talked-to-it/7631 "2026-10-14T18:13:49Z")
Researchers tracked down a rogue fleet of Chinese AI agents operating in the wild. They then talked to it. The encounter shows it is possible to interact with an autonomous agent fleet outside controlled settings. A Fle… --- ## [OpenAI will reportedly ship at least one model this year but its flagship still needs alignment work](https://forum.gnoppix.org/t/openai-will-reportedly-ship-at-least-one-model-this-year-but-its-flagship-still-needs-alignment-work/7630)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 4:24pm UTC](https://forum.gnoppix.org/t/openai-will-reportedly-ship-at-least-one-model-this-year-but-its-flagship-still-needs-alignment-work/7630 "2026-10-14T16:24:28Z")
OpenAI will ship at least one new AI model this year, but its flagship model is still not ready for public release because it requires more alignment work. That is according to recent reports about the company’s internal… --- ## [Nvidia's Jensen Huang pledges $2 billion for Europe, $250 million in compute for UK safety labs](https://forum.gnoppix.org/t/nvidias-jensen-huang-pledges-2-billion-for-europe-250-million-in-compute-for-uk-safety-labs/7629)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 2:37pm UTC](https://forum.gnoppix.org/t/nvidias-jensen-huang-pledges-2-billion-for-europe-250-million-in-compute-for-uk-safety-labs/7629 "2026-10-14T14:37:56Z")
Nvidia pledges $2 billion for Europe, with $250 million in compute for UK safety labs Nvidia CEO Jensen Huang has pledged $2 billion for Europe, with $250 million in compute reserved for UK safety labs. The commitment pa… --- ## [AI guardrails designed to stop hackers are now blocking defenders, and Orion-1 aims to fill the gap](https://forum.gnoppix.org/t/ai-guardrails-designed-to-stop-hackers-are-now-blocking-defenders-and-orion-1-aims-to-fill-the-gap/7628)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 1:37pm UTC](https://forum.gnoppix.org/t/ai-guardrails-designed-to-stop-hackers-are-now-blocking-defenders-and-orion-1-aims-to-fill-the-gap/7628 "2026-10-14T13:37:59Z")
AI safety guardrails built to block hackers are now stopping the security defenders they were meant to protect. Cybersecurity teams are finding that standard AI filters refuse to help with essential defensive tasks like … --- ## [Alleged OpenAI spy vibe-coded a surveillance device using his wife's chatbot account](https://forum.gnoppix.org/t/alleged-openai-spy-vibe-coded-a-surveillance-device-using-his-wifes-chatbot-account/7627)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 12:53pm UTC](https://forum.gnoppix.org/t/alleged-openai-spy-vibe-coded-a-surveillance-device-using-his-wifes-chatbot-account/7627 "2026-10-14T12:53:06Z")
An alleged OpenAI spy used his wife’s chatbot account to vibe code a surveillance device, according to available details. The timing and motive are still unclear, but the method is unusually specific. The accusation rem… --- ## [Gnoppix Security Announcement - 2026-10-14T04:38:07.536-04:00](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-14t0407-536-04-00/7626)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 14, 2026, 8:38am UTC](https://forum.gnoppix.org/t/gnoppix-security-announcement-2026-10-14t0407-536-04-00/7626 "2026-10-14T08:38:08Z")
GSA-6558-1: curl Security Update Gnoppix Security issued GSA-6558-1 in 2026 to address vulnerabilities in curl. The update affects Gnoppix 24/25 (oldstable) and Gnoppix 26 (stable). Users should install the fixed packa… --- ## [406 scientists demand AI models be cut off from the internet immediately](https://forum.gnoppix.org/t/406-scientists-demand-ai-models-be-cut-off-from-the-internet-immediately/7625)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 7:02pm UTC](https://forum.gnoppix.org/t/406-scientists-demand-ai-models-be-cut-off-from-the-internet-immediately/7625 "2026-10-13T19:02:41Z")
Scientists Demand Immediate Internet Cutoff for All AI Models 406 scientists have signed an open letter demanding that all AI models be cut off from the internet immediately. The letter represents a direct and time-sensi… --- ## [Gemini 4 Argon is now available worldwide as Google's most capable frontier model](https://forum.gnoppix.org/t/gemini-4-argon-is-now-available-worldwide-as-googles-most-capable-frontier-model/7624)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 6:15pm UTC](https://forum.gnoppix.org/t/gemini-4-argon-is-now-available-worldwide-as-googles-most-capable-frontier-model/7624 "2026-10-13T18:15:24Z")
Gemini 4 Argon is live worldwide as Google’s most capable frontier model Google has released Gemini 4 Argon, its most capable frontier model, to users worldwide. The launch is effective immediately. Argon is part of the… --- ## [Machine learning pioneer Michael I. Jordan dismantles AI doom rhetoric as "grand moralizing performance"](https://forum.gnoppix.org/t/machine-learning-pioneer-michael-i-jordan-dismantles-ai-doom-rhetoric-as-grand-moralizing-performance/7623)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 5:24pm UTC](https://forum.gnoppix.org/t/machine-learning-pioneer-michael-i-jordan-dismantles-ai-doom-rhetoric-as-grand-moralizing-performance/7623 "2026-10-13T17:24:21Z")
Michael I. Jordan, a pioneering figure in machine learning, has publicly dismantled the pervasive AI doom rhetoric, calling it a “grand moralizing performance.” The statement cuts through the fearmongering that has domin… --- ## [Character.ai will shut down if no buyer emerges, capping a fall from $2.7 billion Google deal to bankruptcy](https://forum.gnoppix.org/t/character-ai-will-shut-down-if-no-buyer-emerges-capping-a-fall-from-2-7-billion-google-deal-to-bankruptcy/7622)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 3:36pm UTC](https://forum.gnoppix.org/t/character-ai-will-shut-down-if-no-buyer-emerges-capping-a-fall-from-2-7-billion-google-deal-to-bankruptcy/7622 "2026-10-13T15:36:29Z")
Character.AI will shut down if no buyer emerges, capping a stunning fall from a $2.7 billion Google deal to the brink of bankruptcy. The company, once a leading force in conversational AI, now faces an existential crisis… --- ## [Western mathematicians fear AI makes them dumber while China builds an entire alliance for it](https://forum.gnoppix.org/t/western-mathematicians-fear-ai-makes-them-dumber-while-china-builds-an-entire-alliance-for-it/7621)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 2:50pm UTC](https://forum.gnoppix.org/t/western-mathematicians-fear-ai-makes-them-dumber-while-china-builds-an-entire-alliance-for-it/7621 "2026-10-13T14:50:33Z")
The Lede Western mathematicians fear that relying on AI in research makes them dumber. China is building an entire alliance to push AI into mathematics instead. That split captures one of the most important debates in th… --- ## [Anthropic says it cut Claude reasoning token theft by up to 99% in fight against industrial-scale distillation](https://forum.gnoppix.org/t/anthropic-says-it-cut-claude-reasoning-token-theft-by-up-to-99-in-fight-against-industrial-scale-distillation/7620)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 11:12am UTC](https://forum.gnoppix.org/t/anthropic-says-it-cut-claude-reasoning-token-theft-by-up-to-99-in-fight-against-industrial-scale-distillation/7620 "2026-10-13T11:12:40Z")
Anthropic Cuts Claude Reasoning Token Theft by Up to 99% Anthropic says it has cut Claude’s reasoning token theft by up to 99% as part of an effort to stop “industrial-scale distillation” of its AI models. The company h… --- ## [Anthropic study warns that small AI models could soon replicate across everyday devices](https://forum.gnoppix.org/t/anthropic-study-warns-that-small-ai-models-could-soon-replicate-across-everyday-devices/7619)
**Author:** [@amu](https://forum.gnoppix.org/u/amu)\ **Replies:** 0\ **Last updated:** [October 13, 2026, 10:37am UTC](https://forum.gnoppix.org/t/anthropic-study-warns-that-small-ai-models-could-soon-replicate-across-everyday-devices/7619 "2026-10-13T10:37:43Z")
Anthropic Study Warns Small AI Models Could Soon Replicate on Everyday Devices A new Anthropic study warns that small AI models could soon replicate themselves across everyday devices. The threat is no longer limited to … [Previous page](https://forum.gnoppix.org/latest.md?page=1) [Next page](https://forum.gnoppix.org/latest.md?page=3)