THE DAILY EDITION

AI

8 items · 19 articles in this topic

Google and Microsoft both announced new model families, and Cloudflare opened weights for its extraction models. Privacy and misclassification harms from deployed AI also made the news.

  1. 1

    Google announced Gemini 4 Neon, a new Gemini model. A blog post argues it looks like a competitor to Anthropic's Opus 5.5.

    HighHacker News (front page) ↗8 min read
  2. 2

    Google announced Genie 4, its world model, is now available through Project Genie.

    HighHacker News (front page) ↗6 min read
  3. 3

    Microsoft announced seven new in-house AI models at Ignite 2026, stressing price-performance.

    HighHacker News (front page) ↗5 min read
  4. 4

    Cloudflare graduated Clairic from preview and released Clairic and Clairic-flash as open-weight extraction models on Workers AI.

    NotableCloudflare Blog ↗5 min read
  5. 5

    The Verge reports that the frontier AI pause discussed earlier in the year has faded, as Nvidia ships Vera Rubin NVL72 systems.

    HighThe Verge ↗6 min read
  6. 6

    Internal Meta chats show staff alarmed by Muse AI outputs showing random users' bank and bitcoin balances. Meta says the figures are not real.

    NotableArs Technica ↗7 min read
  7. 7

    Black high school students sued after an AI gun-detection system flagged a Doritos bag as a firearm, leading to detention and searches.

    NotableHacker News (front page) ↗4 min read
  8. 8

    The Register reports that uncensored AI models are widely distributed and used on public code hosting platforms.

    NotableThe Register ↗7 min read
More in AI · 11 articles›
  • Canva adds a toggle to disable its generative AI features, while background removal and Magic Grab stay on.

    The Vergecanvagenerative-aiconsumer-app
    Feed excerpt

    Canva will now allow you to turn off its generative AI features. With this change, you can disable the graphic design platform's array of AI tools, such as its AI photo-extending Magic Expand feature, as well as turn off the ability to generate images with its Canva chatbot. This toggle doesn't turn off every AI-powered feature on Canva, however. That includes Canva's background remover and its Magic Grab tool, which lets you select an object in an image and move it somewhere else. Canva says these tools will stay on because they "make everyday tasks faster without generating anything new." The toggle still turns off many of Canva's other … Read the full story at The Verge.

  • An IEEE Spectrum article argues that the AI superintelligence debate has serious flaws in its reasoning.

    Hacker News (front page)ai-policysuperintelligence
    Feed excerpt

    Article URL: https://spectrum.ieee.org/ai-superintelligence-debate Comments URL: https://news.ycombinator.com/item?id=50559140 Points: 69 # Comments: 71

  • Two unidentified stealth AI models appeared on a leaderboard, both ranking above Fable 5.2.

    Hacker News (front page)llmbenchmarks
    Feed excerpt

    Article URL: https://contralabs.com/frontierdesign Comments URL: https://news.ycombinator.com/item?id=50558559 Points: 33 # Comments: 2

  • Simulates whether LLM-issued action validation by decision models improves automated network discovery across network scales.

    arXiv cs.CR (Cryptography & Security)llmpentestingarxiv
    Feed excerpt

    arXiv:2611.23947v1 Announce Type: new Abstract: Network discovery, the identification of live hosts and their exposed services, is the first phase of every penetration test, and it is increasingly automated by harnesses driven by large language models (LLMs). To mitigate hallucinated or unsafe commands, recent work proposes inserting fast, non-generative decision models to validate actions prior to execution. This paper quantifies whether such validation improves discovery across three network scales (n = 10, 100, and 1,000) in a simulation calibrated on a 30-node containerized testbed. Our findings show that chunked batch dispatch recovers 92% to 100% of open ports in large-scale networks, whereas single-host dispatch collapses to under 6% due to context saturation and scan latency. Furthermore, at the measured proposal quality of the local model, decision gating increases its port recall from 75.5% to 89.0% on small networks and intercepts costly decoy tarpits, but it imposes additional latency on frontier models during early host discovery.

  • A blog post argues that the unreleased Gemini 4 Neon model looks like a competitor to Anthropic's Opus 5.5.

    Hacker News (front page)llmgeminimodel-release
    Feed excerpt

    Article URL: https://micjm.com/blog/gemini-4-neon-looks-like-a-top-model/ Comments URL: https://news.ycombinator.com/item?id=50554374 Points: 107 # Comments: 36

  • Satya Nadella says AI agents could disrupt advertising-driven web business models and reshape competition, including for Amazon.

    TechCrunchai-agentsmicrosoftadvertising
    Feed excerpt

    Microsoft's CEO says AI agents could disrupt advertising-driven business models, forcing companies like Amazon to rethink how they compete for customers.

  • A blog post describes using LLMs to assist with reverse engineering binaries.

    Hacker News (front page)llmreverse-engineering
    Feed excerpt

    Article URL: https://www.slowchronicle.com/posts/llm-assisted-reverse-engineering Comments URL: https://news.ycombinator.com/item?id=50553225 Points: 15 # Comments: 0

  • Google has made its Gemini Tasks personal assistant feature free for users, with new features announced on its blog.

    Hacker News (front page)geminigoogleassistant
    Feed excerpt

    Article URL: https://blog.google/innovation-and-ai/products/gemini-app/tasks-in-gemini-free-new-features/ Comments URL: https://news.ycombinator.com/item?id=50551601 Points: 11 # Comments: 4

  • Snapchat lets developers build AR Lenses by describing them in plain language to AI assistants like Claude Code and Cursor.

    TechCrunchsnapchatardeveloper-tools
    Feed excerpt

    Lens creators can describe the Lens they want in plain language, and their AI assistant handles the process from start to finish, Snapchat says.

  • Google added temporary threads to AI Mode in Search, letting users chat without saving history or influencing recommendations.

    TechCrunchgoogleai-modesearchprivacy
    Feed excerpt

    Google's new temporary threads let users have conversations in AI Mode without adding them to their search history or influencing future recommendations.

  • OpenAI published details on using its models to advance chip design.

    Hacker News (front page)llmhardwareopenai
    Feed excerpt

    Article URL: https://openai.com/index/advancing-chip-design-with-our-models/ Comments URL: https://news.ycombinator.com/item?id=50549493 Points: 82 # Comments: 44

Written by claude-sonnet-5-5 from feed titles and summaries. The model can misstate details; every claim links to its source.