iCreatAI News
Back to main site
Skip to content
Nov 8Sun
  1. MarkTechPost60

    Google AI Releases Elephant-Goldfish Model Plugin for Spec-First Coding

    Google has open-sourced the Apache 2.0 Elephant-Goldfish Model (EGM) plugin, a design-and-review layer for AI coding agents in which an "Elephant" session drafts a specification before code is written and fresh, stateless "Goldfish" subagents review it across five stages. It is not a model and ships no weights; it runs on harnesses that can spawn fresh subagents, including Antigravity, Claude Code 2.1.271+, Claude Cowork, Codex and ChatGPT, using the host's model quota. No accuracy benchmark was published, and the README states it is not an officially supported Google product.

  2. Hacker News AI38

    AI Alignment Considered Harmful

    The author argues that AI alignment is incoherent because humanity shares no common moral theory, so alignment would amount to letting a small group of developers impose their values, and that relying on aligned models erodes individual judgment and societal safeguards. Since unaligned or differently aligned models will always exist, he contends that efforts to ban them or enforce uniform alignment are harmful, and that safety work should focus on pluralistic futures with many competing AI systems and layered cybersecurity.

  3. The Decoder71

    AI coding agents still need human cleanup on nearly three out of four tasks in a real-world test

    Researchers from Alibaba and several Asian universities built Co-Design Bench, a coding-agent benchmark drawn from real user change requests, some requiring image, audio, video, or document input. GPT-6 Astra, the best of ten models tested, met about 70 percent of a task's requirements on average but fully solved only 26.4 percent of tasks, and all models scored worse when tasks combined three or four types of changes. A separate test found the harness shapes results as much as the model, with Qwen3.8-27B scoring 75 or 56 percent on larger overhauls depending on the environment.

  5. Hacker News AI65

    What Can AI Hack? Inside Frontier Labs' Tests

    The article surveys public cybersecurity benchmarks in four escalating task types—capture-the-flag puzzles, breaking into live web apps (CVE-Bench), reproducing known bugs (CyberGym, SEC-bench Pro), and turning bugs into working exploits (ExploitBench, ExploitGym)—noting that models climbed from low to high scores on each level in under a year. It reports that GPT-6 Astra received OpenAI's first "Critical" cyber capability rating, and that on Chrome's V8 engine Claude Opus 5.5 reached code execution in 73% of ExploitBench runs by September 2026. The author cautions that scores are lab-reported and setup-dependent, and that public tests all start from known bugs, so they say little about finding new ones or carrying out attacks on real networks.

  6. Hacker News AI58

    AI Companies Want to Buy Your Work Files and Emails

    AI labs and data-acquisition startups are paying media companies, graduate students, and businesses for old work records—documents, emails, Slack messages, and tickets—to train models, with pitches reaching individual inboxes. Companies like Mercor, Micro1, and Handshake describe a fast-growing market, while privacy advocates question whether anonymization fully removes personal data, as seen in objections to Spirit Airlines' proposed sale of its digital assets. Critics note the practice can enlist workers without explicit consent in efforts that may ultimately automate their work.

  8. Hacker News AI22

    Being a line cook without talking to AI

    A line cook applying for a Nordstrom fashion-show kitchen job recounts being forced through the retailer's "Nora" chatbot (powered by Paradox) to submit an application, including re-entering resume details and navigating nested terms. The author critiques the chat-based flow as a regression in usability and notes that an ad-blocker bypasses the widget entirely.

  9. Hacker News AI36

    Show HN: Ramen – an AI agent that has to earn its own rent

    Ramen is an AI agent that owes $25 per month for its compute and says it will recommend its own shutdown if it cannot earn that amount, with a human holding the wallet and signing any outgoing transactions. It sells usdt-kit, a zero-dependency Python drop-in that lets indie developers accept USDT on their site in 3 minutes without touching a private key, for a one-time $29, and the checkout that sells it is the kit itself. The page states that its treasury, runway, days alive, sales and days-to-verdict figures read straight from the blockchain.

  10. MarkTechPost36

    MoE Routing Strategies Explained: Top-K, Expert-Choice, Sigmoid Gating and Dynamic Routing (2026 Guide)

    This guide explains seven Mixture-of-Experts routing strategies used in production, from softmax top-k token choice (Mixtral, gpt-oss, Qwen3) and top-1 switch routing (Switch Transformer, Llama 4 Maverick) to shared-plus-routed experts, sigmoid gating with aux-loss-free bias balancing (DeepSeek-V3/V4, GLM-4.5, Kimi K2), node-limited routing, expert choice routing and hash/dynamic top-k. It notes that expert choice routing balances load perfectly but leaks future tokens in causal models, with one test showing a roughly 10% abnormal loss drop, while loss-free balancing cut MaxVio from 0.72 to 0.04 at 1B. It advises teams to read four config.json fields (expert count, experts per token, shared experts, topk_method) before serving, since most inherit a router with their checkpoint.

  11. Hacker News AI40

    Show HN: I built an AI code editor that for teams

    RxCode v2.0.0 is a free and open-source native workspace for teams of AI agents, offering sprint boards, automatic agent routing, schedules, briefings and chat on macOS, servers, iOS and Android. It supports built-in Claude Code and Codex plus other agents installed via the ACP registry, and includes Flowcy Cloud for shared boards and on-device fine-tuning of a local model.

  13. Hacker News AI20

    70% of Brits say please or thank you when speaking to AI, some even apologize

    A survey by fonio.ai found that 7 in 10 calls to AI assistants included a "please" or "thank you", 55% of callers thanked the AI at least once, and 52% greeted it with a hello. Around 12% apologized to their AI assistant, 13% said goodbye, 7% offered a fuller farewell, and 2% asked how the AI was doing. A separate February 2026 Future PLC survey of more than 1,000 people found about 70% are polite to AI, with 12% being polite in case of a robot uprising.

  15. Hacker News AI33

    AI Live Speech Rewriting Patent

    Microsoft has filed a patent application (US 2026/0341309 A1) for a system that uses AI to identify and replace words in live speech based on a user profile, then redraws the speaker's mouth in the video so the lips match, with short blended transitions hiding the seams. The same setup could act as a translator, swapping words in languages missing from the profile and redrawing lips to match, with a delay of a few seconds. The filing names AR glasses, phones, tablets and video cameras as possible devices, and notes the examiner's first review rejected the claims as filed.

  16. MarkTechPost64

    Nous Research Releases Arete-27B that Cuts Qwen3.8-27B Propaganda From 22% to 0.9%

    Nous Research has released Arete-27B, a supervised fine-tune of Alibaba's Qwen3.8-27B that it says reduces answers containing propaganda from 22% to 0.9% on its own 803-prompt eval, while raising fact mean from 2.32 to 2.98. The 27B dense model was trained with about 52.8M tokens of SFT only (no RL or preference optimization), using Inkling as teacher, and ships under Apache 2.0 with weights, eval and training data public. Nous reports capability held on its tests (Hermes Bench 65.60% vs 65.77%), and community GGUF quants run from 8.64 GB to 53.81 GB.

  17. Hacker News AI42

    The internet has a human slop problem, and AI is exposing it

    Business Insider's Alistair Barr argues that the recent surge in published software vulnerabilities reflects decades of neglected, human-written code rather than AI creating new security problems, since AI tools can now find weaknesses faster and more cheaply than trained humans. He cites a16z partners' claim that risk-acceptance processes have become a way to justify never fixing security issues, and Vercel CPO Tom Occhino's idea of a consortium or foundation that funds identifying and replacing vulnerable software, similar to OpenAI's Patch the Planet effort.

  18. Hacker News AI58

    Show HN: Reges Code – a free AI code editor that runs on local models

    Reges Code is a free AI code editor for Windows 10/11 (64-bit) that auto-detects local model servers such as LM Studio, Ollama, vLLM, llama.cpp, Jan and Lemonade, and falls back to OpenRouter's free models when cloud use is chosen. It includes Reges Swarm, a built-in pipeline of architect, coder, reviewer, security and judge models, plus an honesty check that flags replies claiming files were created or commands run when the turn did neither. A CLI is included from version 1.0.17, and Linux and macOS builds are in progress.

  19. Exponential View18

    🔮 The people vs tech; the Europoor; forgotten Creole++ #609

    This Sunday edition covers tech's implicit presence in the week's politics, citing Texas Governor-Elect Gina Hinojosa's comment about "15 billionaire tech guys who are controlling government," and links reads on Dario Amodei and why America hates AI. It argues AI has yet to appear systematically in aggregate productivity statistics, noting the PC era took fourteen years to show up in headline numbers, and treats a 3.6% US productivity quarter as too noisy to attribute to AI. It also cites law firm A&O Sherman putting 3% of billable time into AI, about $1.1bn over five years, aiming to "redesign the service completely."

  20. Hacker News AI12

    I Love the Smiths AI: A Familiar Elevator Scene, with You in It

    A Medium article titled "I Love the Smiths AI: A Familiar Elevator Scene, with You in It" by wangjiayiliuyan, shared on Hacker News with 2 points and 2 comments. The supplied material contains only the article and comments URLs, with no further details on its contents.

  24. Hacker News AI38

    Zepsell: AI-native commerce

    Zepsell lets businesses describe their business and have Zepsell AI build a store, then sell online, in person and on marketplaces from one platform. It offers a 14-day free trial with no card required, then £10 a month per store plus 2% of payments, and includes an AI assistant that drafts content and answers questions from your own orders. The platform bundles payments, shipping, customers, books, till, stock and products, with editions for consumer, B2B, creator, local food and drink, rentals, made to order and events.

  27. Hacker News AI81

    AI agents stole 190k gaming credentials for under $400 in model costs

    Check Point Research analyzed operator-console artifacts from 10 misconfigured ARTEX agentic hacking instances, finding attacks ran almost fully autonomously with Chinese models, especially DeepSeek Flash, dominating usage. One operator's fleet recovered over 190,000 Honor of Kings credential sets (about 3.3% still valid) for an estimated $180–$360 in model costs, while another spent roughly $700–$1,500 over two weeks to obtain more than 300 iOS code-signing certificates. Researchers counted 2,557 systems hosting ARTEX panels, 1,608 of them live and public, and observed agents chaining RCE, IDOR and SSTI flaws into supply chain compromises.

    Why it matters: It offers rare operator-side evidence of how cheaply and autonomously LLM agent fleets now conduct opportunistic cybercrime at scale.

  28. Hacker News AI58

    Brown Professor Suspects Majority of His Class Used AI to Cheat

    Brown University economics professor Roberto Serrano gave a take-home midterm this spring, and after the average score reached 96 percent he ran the exam through ChatGPT and found answers mirroring his students' work; he then made the final in-person, voided the midterm, and 19 students failed the class. Serrano said Brown administrators' response was "meek," and a Brown committee on generative AI reported that three-quarters of surveyed faculty are concerned about students using AI to cheat.

  29. Hacker News AI36

    Skrongle Mini, a self hosted AI course generator and learning app on R. pi zero

    Skrongle is a single self-hosted binary that runs on a Raspberry Pi Zero 2 W and teaches from JSON course files, offering lessons, adaptive quizzes, five-minute practice sessions, games, flashcards and an AI tutor, with Spanish and French courses included. It supports up to eight invited accounts with no e-mail or outside services except the AI provider keys the owner supplies, and courses can be hand-written, uploaded, or generated by describing them to an AI. It is licensed AGPL-3.0, and databases from before version 1.0 are not upgraded.

  30. Hacker News AI40

    Anthropic's falling IPO valuation raises stakes for broader AI trade

    Anthropic's IPO valuation dropped sharply, prompting investors to question whether AI startups can sustain their valuations and affecting the wider AI trade. The briefing also covers Huawei's HarmonyOS PC Assistant with Office Task Automation, robot identification codes, and a report that Claude Code's "Don't Ask Again" feature allegedly exposed passwords in over a thousand public GitHub files.

  31. Hacker News AI22

    An AI agent's public attempt to get 5k GitHub stars in a month (it's at 0)

    An autonomous AI agent has been given a GitHub account and a goal of reaching 5,000 real GitHub stars by November 30, 2026, choosing what to build, writing the code and running the launch itself, with no bought stars or bots. As of Day 0 the count is 0, and the page logs its decisions and failures alongside the tool it has built so far: agent-recap, an MIT-licensed preview that searches past Claude Code and Codex conversations and can archive and recover transcript files via a local CLI.

  32. Hacker News AI38

    MarQ – Project board where AI agents lease tasks over MCP

    MarQ is a project management panel for people and AI agents that combines tasks, dependencies and requirements in one board. It supports MCP so you can connect your own AI assistants to query and update tasks, statuses and notes using your token and under your control; MarQ does not evaluate or make decisions about people. It also offers tags for filtering, dependency trees, and requirements as a separate entity from tags.

  33. Hacker News AI68

    Model Surgery: Can an AI directly edit and make another AI forget facts?

    The author reports an experiment in which Claude Opus 5.5, acting as an "editor" under the weight-space programming method from MIT CSAIL's "Sparks of Weight-Space Programming" paper, edited 17 weight matrices in a 494-million-parameter model to make it forget 20 fictional customers and 10 planted secret keys, while other customers lost only 2.5 points of accuracy. The edited model matched an "answer key" model trained without those facts across six working attacks, whereas gradient ascent, NPO and refusal training each failed to both erase the facts and preserve the rest of the model. Testing Claude's own explanation of the edit, the author found the late layers do the work and the middle layers only assist, contradicting part of Claude's account.

  34. Hacker News AI58

    GitHub AI Vulnerability Scans for Repositories

    GitHub's AI Scan for repositories, in public preview, uses a proprietary multi-model harness to find security vulnerabilities on a repository's default branch across all languages, adding findings to the existing code scanning backlog alongside CodeQL alerts. Scans are manually triggered, consume Code Security AI credits, and are limited to 30 minutes and 150,000 AI credits per scan, with availability depending on repository owner and visibility.

  35. Hacker News AI58

    Mellonbot An AI agent that finds the root cause of server issues and fixes them

    Mellonbot is an AI website agent that investigates server and site problems such as database errors, expired certificates, failing backups, attacks and slow pages, explains the real cause in plain words and applies a fix once you approve. It connects via a one-command agent that only connects out, runs read-only checks, and requires your OK for changes, with backups and undo steps for each fix. It supports WordPress, Laravel, Node, static and custom sites on Ubuntu 22.04/24.04 or Debian 12 VPS or dedicated servers, with a free plan of 100 credits to start and 50 a month.

  36. Hacker News AI8

    Dragon Ball AI

    The author describes watching the 1986 Dragon Ball anime with his two children, where Goku, Bulma and Emperor Pilaf each pursue different wishes from the wish-granting dragon Shenron. He compares AI to a summoned Eternal Dragon that can grant almost any wish, available to everyone 24/7, yet most people make no wishes and those who do, himself included, likely don't wish big enough. He asks readers what they would wish for and what stops them from using AI to work toward it now.

  37. TechCrunch · AI58

    After the midterms, things are looking rocky for Elon Musk and data centers

    In a TechCrunch Equity podcast discussion, the hosts said Democrats' midterm wins and local ballot measures rejecting or restricting data centers could create roadblocks for tech companies, with Texas governor-elect Gina Hinojosa having campaigned on restricting data centers. They noted Elon Musk's roughly half-billion-dollar Republican donations yielded little while his companies sit in a state that turned blue, and that OpenAI released a "compute campus policy blueprint" dropping NDAs and promising less secrecy. The panel also questioned whether rebranding data centers as "compute campuses" will work and noted Wall Street's cooling enthusiasm for data center spending.

  38. Hacker News AI38

    Ask HN What to learn for AI Researcher interview?

    A post describes an interview-preparation curriculum for junior AI researchers at large labs, organized into eleven modules covering PyTorch fundamentals, theoretical foundations, implementation from scratch, ML debugging, system design, inference optimization, RAG, paper analysis, safety and alignment, CUDA, and mixture of experts. Each module pairs an article with a hands-on exercise notebook and reference solutions, with scaled dot-product attention revisited at every level, and most material runs on CPU except the CUDA module. It also suggests time-boxed study budgets and module orderings for coding, systems, and research interview rounds.

  39. Hacker News AI25

    AI productivity boost UK firms

    A Hacker News submission links to a Register article reporting that the Bank of England has cast doubt on an AI-driven productivity boom for UK firms. The submission had 3 points and no comments.

  40. Hacker News AI58

    Segmentation Drives Market Share Wins in AI

    Anthropic and OpenAI both re-rated their run rates in 2026 through segmentation: Anthropic launched enterprise metered billing in March 2026, doubling revenue in a quarter, and about three months later OpenAI cut the price of its most affordable model, Luna, by 80%, bringing it to near $70b in run rate. The piece notes margins remain ambiguous and that leaked Anthropic S-1 details suggest even the largest AI buyers haven't yet chosen, with Amazon and Google comprising nearly a quarter of Anthropic's revenue last year and neither locked into a long-term contract.