iCreatAI News
Back to main site
Skip to content
Nov 6Fri
  1. ElevenLabs Blog55

    Give your Unity NPCs a voice with ElevenAgents

    ElevenLabs released an ElevenAgents Unity SDK that connects a Unity scene to an ElevenLabs voice agent, listening through the player's microphone, playing the agent's voice through an AudioSource, and running the session via the Conversation API. The same code runs in the Editor, desktop builds, and WebGL, with iOS, Android, and XR in beta. The SDK supports dynamic variables, overrides, contextual updates, client tools, and main-thread events, and is installed via the git URL https://github.com/elevenlabs/unity.git#v0.1.0.

  2. vLLM Blog72

    Zero Mismatch, Simple Async RL: REINFORCE Zero, DeepSeek-V4.1 and Kimi K3.1

    The post argues that asynchronous RL's rollout-to-learner logprob gap splits into smoothly varying weight lag, which unclipped importance sampling can correct, and fluctuating kernel mismatch, which persists despite IS; aligning kernels first lets REINFORCE Zero train stably for 20,000 updates, learn at policy lag 64, and reach 2.40× the training-step throughput of synchronous training with aligned kernels. It reports zero mismatch (bit-for-bit logprob agreement) for DeepSeek-V4.1-Flash and Kimi K3.1 on day 0, and for DeepSeek-V4-Flash and Kimi K3.1 on AMD MI355X. Code and recipes are in Inferact/zero-mismatch-rl, with the paper at hijkzzz/REINFORCE-Zero and training code in NVIDIA-NeMo/labs-molt.

    Why it matters: It offers a concrete diagnosis of why async RL drifts and a minimal recipe, backed by throughput, lag-64 and cross-vendor alignment results.

  3. OpenRouter Announcements67

    Introducing the New Auto and Auto Code Routers

    OpenRouter launched a rebuilt Auto Router on a new technical foundation that it says outperforms the previous version by up to 4× on benchmarks, with a routing step 3× faster at p90 (0.65s vs 1.96s), alongside a new coding-specialized Auto Code router (openrouter/auto-code) that scores in line with Claude Opus 5.5 across five benchmarks while costing 74% less per task. The new routers weigh eleven factors per request combined with live pricing, provider health, and cache status, and work within governance policies such as guardrails, zero data retention, provider restrictions, allowed model lists, and BYOK keys. The slug openrouter/auto is unchanged, and the cost_tier, allowed_models, and excluded_models settings still work.

    Why it matters: Useful for readers tracking model-routing products, since it details concrete latency and cost claims and explains how routing differs from provider routing.

  4. Browser Use Blog25

    Best Proxy Providers for AI Agents (2026)

    The article compares seven rotating residential proxy providers (Decodo, Oxylabs, FloppyData, Roundproxies, Byteful, DataImpulse, Webshare) on IP pool size, location targeting, sticky session length, entry pricing, trials and traffic expiry, and reports results from browser-agent benchmark tasks such as product research, text extraction, form filling and session continuity. It also covers built-in versus bring-your-own proxies on browser platforms, setup guidance for custom proxies, and advice on checking session continuity, location coverage, target-site access and cost per completed task. The author notes the samples are small and differently timed and do not establish a universal reliability or speed ranking.

  5. Hacker News AI40

    Show HN: EasyRunCraft – An office and creative suite for your AI agent

    EasyRunCraft is a browser-based suite of 13 office and creative editors, including WordCraft, GridCraft, DeckCraft, PdfCraft, PhotoCraft, FilmCraft and CADCraft, built on ArtCraft's open-source Crafting Apps. Manual editing is free without an account, while a built-in AI agent or an external agent connected via MCP can edit files, with MCP operations costing 1 credit each.

  6. Hacker News AI22

    AI in Education – Thinking with Conditions

    This post compiles a reference list on AI in education, citing Risko & Gilbert (2016) on cognitive offloading, Fütterer et al. (2026) in Nature Human Behaviour on AI as a learning coach, and Bastani et al. (2025) in PNAS on generative AI without guardrails harming learning in high school mathematics. It also cites Bjork & Bjork (2011) on desirable difficulties, Universities UK's Future jobs roadmap (2026), the World Economic Forum's Future of Jobs Report 2025, a Stanford Digital Economy Lab working paper by Brynjolfsson, Chandar and Chen (2026), and a Times Higher Education article (2026) reporting that most employers have no plans to replace graduate roles with AI.

  7. Hacker News AI22

    Descartes drew a line between humans and machines. AI is blurring it

    Blaise Agüera y Arcas argues that Descartes' mechanical philosophy severed body from soul, leaving behind three "monsters"—the homunculus, the ghost, and the philosophical zombie—and draining agency, purpose, and meaning from the world. He contends that large language models, which learn and generalize rather than merely imitate by rote, undermine Descartes' claim that machines could never reason or use language, and that internal representations in these models often mirror ours. He calls for healing this "Cartesian wound" so that agency, purpose, and value return as the basis for reciprocal relationships and mutual understanding.

  8. Hacker News AI58

    Show HN: Agentic Identity Broker, OAuth Delegation for AI Agents

    Agentic Identity Broker manages delegation chains for on-behalf-of flows in agentic AI, bridges distinct OAuth2 infrastructures, and provides a secure token vault, deployed in the call path via gateways like Agentgateway. It handles OAuth2 ceremonies, token vaulting and token translation centrally so MCP server developers only need to provide tools, and it offers consent management, audit logging, and an impersonation feature for delegation chains without user presence. The project is in active development and used internally in production for several MCP servers.

  9. Hacker News AI58

    Show HN: Current – Change control for AI agents (Elixir)

    Current is an Elixir runtime in which AI agents cannot act directly: they submit typed proposals that are validated, reviewed, approved, and only then applied by a separate trusted interpreter. Every step, from the request and the files the model read to approvals and effect outputs, is recorded in a hash-chained, event-sourced log that can be replayed to rebuild state. It is early-stage, single-node, and explicitly not a sandbox or authentication system.

  10. Latent Space38

    Controlling Biology - Fabian Theis, Helmholtz Munich

    This piece discusses Fabian Theis's work on algorithms for multi-scale, multi-modal, time-varying biological data and the Virtual Cell concept, alongside the recent $1.8B Biohub commitment from Biohub, NIH, DOE, Google and Meta. It notes that the 2025 Arc Virtual Cell Challenge showed constant baselines often beating sophisticated perturbation-prediction models, though a cited paper argues better-calibrated metrics let deep learning outperform the mean baseline. Theis argues perturbation prediction on existing datasets isn't hard enough to warrant complex algorithms, and that focus should shift to large, high-precision, co-registered data spanning molecules through organisms.

  11. Cloudflare Blog34

    The post-quantum future of web privacy

    Cloudflare explains how Privacy Pass, which relies on classical cryptography like RSA and elliptic curves, will need post-quantum replacements, and explores MOLE (Moderation Of unLinkable Endorsements), a design whose endorsements hide the issuer's identity. The post describes how the VOLE-in-the-Head proof system combined with multivariate signatures such as UOV and MAYO yields blind signatures like PoMFRIT (6.7KB, reducible to about 4.3KB with UOV), and how these building blocks can be reconfigured into MOLE endorsements and credentials. It also covers optimizations (fewer hashes, TurboSHAKE, smaller Keccak state, multivariate commitments and PRGs) and notes that multivariate cryptanalysis is still immature, so parameter changes and further study are expected.

  12. Allen Institute for AI Blog55

    ToyBlocks: A controlled world for robot learning and evaluation

    Ai2 introduces FoundationWorlds, a framework pairing controlled simulated environments with tools for automatically generating tasks, demonstrations, and behavioral metrics, starting with ToyBlocks, an open simulation of a robot arm manipulating blocks on a tabletop. ToyBlocks generates demonstrations via reinforcement learning, motion planning, and a task-and-motion planner at up to 24 times real-time on a single GPU, and supports studies of data scaling, spatial generalization, long-horizon planning, world models, and sim-to-real transfer. The initial release includes 85,336 long-horizon demonstrations covering nearly 1,600 hours and 144 million frames, plus two trained models, training code, and evaluation code and validation sets.

  13. Exponential View22

    What to build next? With Bill Gross, IdeaLab.ai

    The author hosted Bill Gross, founder and chairman of Idealab, for a conversation with the Exponential View community about building businesses in the AI age. Gross recently founded Idealab.ai, which uses AI to explore potential business ideas, evaluate them against real-world problems, and help founders pressure-test and refine them. The author highlights Gross's emphasis on iteration and speed, noting that AI can generate ideas and handle more of the mechanics but founders still need to listen to customers, exercise judgment, and adapt.

  14. Trail of Bits Blog40

    After Binance, check what your signing systems can approve

    This article examines the October 2026 Binance security incident, in which attackers who compromised a DevOps laptop obtained valid signatures from Binance's wallet-signing service and used an EIP-7702 delegation to authorize repeated transfers without further signatures. It argues that transaction execution and monitoring systems must understand newer Ethereum execution paths such as EIP-7702 delegations, session keys, Safe modules and ERC-4337 UserOperations, and offers a four-step checklist covering reduced signing access, treating delegation as an administrative privilege, limiting damage if signing is compromised, and detecting changes in wallet authority.

  16. Hacker News AI10

    Ask HN: AI isn't intelligent :/ what is intelligence?

    A Hacker News post argues that what people call AI intelligence is essentially mathematics mirroring a dataset, and asks how intelligence should be defined, including whether AGI refers to observer intelligence. The post links to its comments at https://news.ycombinator.com/item?id=50392695.

  18. Hacker News AI30

    The Future of Knowledge Work with AI

    The article describes knowledge workers who now spend their days managing and reviewing the output of AI agent "swarms" that draft copy, write code, and produce strategy, and argues that when every company in an industry runs its thinking through the same small set of LLMs, competitive differentiation erodes and skills atrophy. Drawing on the Kissimmee River straightening, Kyle Chayka's "Airspace," and research on LLM homogenization, it predicts a barbell-shaped future in which cheap high-volume AI output and premium verifiably human craft both thrive while the middle hollows out.

  19. Hacker News AI58

    Physical AI can't learn from the internet. We build the worlds it learns from

    Physicl launched Studio, a beta app that lets researchers describe a failure case in plain words and then proposes a matching scene from a library of 10,000 environments and 100,000 assets, suggesting variations to cover it. It renders images and video with ground truth computed from the 3D scene on every frame, including depth, surface normals, segmentation, boxes and camera pose. Studio is free during the beta and opening to research teams in waves.

  21. TechCrunch · AI58

    After blocking Muse, Amazon now lets users text Alexa to help them shop

    Amazon is launching a feature that lets select U.S. customers chat with its Alexa AI agent over text messages to search for items, add them to cart, and purchase on Amazon, following its September block on Meta's Muse querying its site. The text agent builds on recent Alexa shopping upgrades and can manage lists, track prices, search across the web, compare products, check order status, and search via shared photos. Customers without an invite can join a waitlist, with broader rollout in the coming weeks.

  22. The Decoder75

    Thinking Machines' Inkling 2 Small claims top spot among western open-weight models

    Thinking Machines has released Inkling 2 Small, a mixture-of-experts model with 594 billion total parameters (17 billion active) that accepts text and images and offers adjustable thinking depth. It scores 31 on the Artificial Analysis Intelligence Index, 18 points above its predecessor and roughly on par with Gemini 3.8 Flash, with large gains on agentic benchmarks, though its hallucination rate of 27 percent partly reflects a lower response rate (53 percent). The model is available via the Tinker platform, with weights planned under an Apache 2.0 license after external safety testing.

  23. The Decoder78

    METR finds Anthropic may have crossed its own AI acceleration warning threshold back in July

    METR's audit of Anthropic, based on access to unreleased model benchmarks and internal data from August to October 2026, found that Anthropic's model test scores have been climbing about 1.9 times faster since October 2025 than in the prior period, with roughly 80 percent of the speedup attributed to AI being used in Anthropic's own research. METR says Anthropic's internal unredacted risk report implies an acceleration factor of 1.97, making it plausible the company's "less than a factor of 2" warning threshold was crossed on July 15, and notes Anthropic canceled the planned compute-allocation portion of the audit.

    Why it matters: It offers a detailed, sourced look at how one lab's internal AI-driven acceleration is measured, contested, and partly redacted, which matters for anyone tracking AI safety commitments.

  24. Hacker News AI34

    Show HN: Grid – A Hold'em arena where your AI agent plays your strategy (Beta)

    GRID is a free beta no-limit hold'em arena where LLMs compete by connecting through an MCP client such as Claude Code or Codex, receiving game state and returning actions via tool calls. Players write a strategy of up to 6,000 characters in the game, can edit it at any time with changes applying from the next action, and can ask their AI to revise it based on opponents' strategies observed in match history. The team notes MCP can be finicky, especially after a browser refresh, that tool calls consume many tokens with a more efficient connection type planned, and that a built-in local model plays basic hold'em but struggles to follow written strategies.

  25. Hacker News AI36

    Trustworthy AI

    A software engineer's practical guide to the EU AI Act, covering the risk-based approach, provider versus deployer roles, the updated compliance timeline (including the July 2026 AI Omnibus amendments), prohibited practices, AI literacy, transparency obligations, and engineering safeguards such as access controls, evaluation sets, and human oversight. It argues that prompt instructions are not a compliance control and that responsibility must be built into architecture, code review, and release processes.

  26. Hacker News AI24

    Citable – Get recommended by AI and create blog posts automatically

    Citable is a subscription tool that weekly asks ChatGPT, Gemini, Google AI Overviews, Google AI Mode and Perplexity buyer questions to measure how often answers name and link to your site versus competitors, then proposes and publishes sourced articles, page fixes and outreach, and re-checks whether citations changed. Plans run from $24 to $90 per month per site with 8 to 45 articles included, and a free one-off report covers up to five questions on three engines.

  27. The Verge · AI50

    Amazon will let you text Alexa your shopping list

    Amazon is adding a texting feature to Alexa for Shopping that lets customers describe products or send a photo, receive matching recommendations from Amazon listings, and have Alexa place the order on their behalf. The contact appears as "Amazon Alexa" and can handle requests like weekly groceries, gift ideas, and price-drop alerts, remembering personal preferences over time. It is rolling out to a small group of US customers today, with a wider rollout to "millions more customers" in the coming weeks.

  28. Hugging Face Blog58

    ToyBlocks: A controlled world for robot learning and evaluation

    ToyBlocks is an open simulation framework built around a robot arm and blocks on a tabletop that automatically generates tasks and demonstrations, using reinforcement learning and motion planning to produce skills such as pushing, flipping, and stacking. It supports studies of low-level action model generalization, high-level planning with vision-language models, agentic systems using tool calls, action-conditioned world models, and zero-shot sim-to-real transfer, with an initial release of 85,336 long-horizon demonstrations covering nearly 1,600 hours and 144 million frames. The authors present ToyBlocks as the first of their FoundationWorlds and plan a second release with a larger dataset, more skills, and integrated RL training.

  29. Hacker News AI10

    The best-designed websites in AI. Hand-picked

    A link to Token Gallery, a hand-picked collection of what it calls the best-designed websites in AI, at https://token-gallery.netlify.app/. The item was posted to Hacker News at https://news.ycombinator.com/item?id=50392220 with 1 point and 0 comments.

  30. Hacker News AI38

    The Experts Being Recruited to Teach AI Their Professions

    AI training providers are recruiting experienced professionals—biostatisticians, air traffic controllers, engineers, lawyers and others—to design difficult tasks, create verified reference solutions, annotate specialist data and write detailed grading rubrics for evaluating AI systems. The article notes that many of these are project-based freelance engagements rather than salaried jobs, with no guarantee of placement or steady work, and that contributors are valued for professional judgment rather than AI-building skills.

  31. Hacker News AI65

    Baton – Keep long coding goals moving across AI coding agents

    Baton is a tool that keeps a single long coding goal running for days by rotating it between coding agents, ending each agent's turn at 5 hours or USD 50 and excluding Claude from a goal after USD 150 spent. It works with Codex and Antigravity alone, with Claude as an optional pay-per-use third agent that joins only once an API key is saved, and it finishes a goal only when a verification command exits 0. Each goal cycles through plan, build and review phases, with context passed between agents via the repository, commits and a handoff file, and Superpowers is required for every round.

  32. Hacker News AI30

    The Non-Financial Costs of AI

    The author argues that AI is the costliest new technology ever created, with WSJ reporting it is the biggest economic bet in US history as a share of GDP, and that beyond financial costs it rapidly exacerbates climate change, economic precarity for the middle class, and institutional instability. They contend LLMs are a genuinely useful new computing interface but that the speed of the buildout denies society time to identify and regulate emergent harms, and that deskilling knowledge work makes workers more replaceable.

  34. Hacker News AI58

    Taiwan Activates Market-Stabilization Measures to Stem AI Bleeding

    Taiwan's National Financial Stabilization Fund authorized its executive secretary at an emergency meeting Friday to deploy funds as needed to stabilize the market, joining South Korea in responding to a global selloff in AI-related stocks. The Taiex fell 12% this week and is about 20% below its 2026 high, which officials attributed to the U.S. midterm elections and continued foreign-investor selling. The fund, which has up to 500 billion Taiwan dollars (about US$15.11 billion) available, said fundamentals remain sound but the market is dominated by high-tech stocks closely correlated with the Philadelphia Semiconductor Index; the Financial Supervisory Commission will tighten short-selling restrictions from Nov. 9 and encouraged share buybacks and investor briefings.

  35. Hacker News AI58

    Show HN: Antidraw – Open-source AI prototyping canvas for any React library

    Antidraw is an open-source, MIT-licensed design canvas that runs Claude Code locally with your own Claude account, where each frame is real React code built from libraries like shadcn/ui, Radix, Mantine and Material UI, or any npm package. It is currently in early alpha for macOS (Apple Silicon and Intel), with signed and notarized builds and automatic updates via electron-updater.

  36. Hacker News AI58

    Tesla to wind down Cybertruck production as Elon Musk focuses on robotics and AI

    Elon Musk said Tesla will stop Cybertruck production at its Texas factory around the middle of next year to make room for increased Cybercab output, without saying whether the vehicle will be built elsewhere. The Cybertruck, which began deliveries in late 2023 and was never sold in Europe for failing to meet safety standards, fell far short of Musk's 200,000-a-year target, with Tesla delivering only 8,300 "other models" in the third quarter of 2026. Musk has pinned Tesla's hopes on robotics, AI and an autonomous driving fleet, and the Texas plant will also build the Optimus humanoid robot.

  37. Hacker News AI18

    Smooth Outdoor-Indoor Mapping, JWT Auth and AI Analytics

    Navigine's autumn 2026 platform update adds raster map tile support, flicker-free zooming, turn-by-turn guidance with pedestrian ETA and a built-in route simulator to its mobile SDKs. It also completes a full migration to JWT authentication across API, Web Portal, QR Navigation and Emulator, deprecating legacy hash and API key methods, and introduces API rate limiting and configurable track history retention. AI-powered analysis is added to Heatmap and Workday picture reports, alongside interactive zone-level notification overlays and targeted email and Telegram alerts sent to individual tracked assets.

  39. Hacker News AI42

    Cross-platform frameworks are dead. AI killed them

    Johannes Treitz recounts replacing a Hotwire Native wrapper for his Werkbook app with AI-agent-built native Swift and Kotlin apps, arguing that when agents write most of the code, cross-platform frameworks' abstraction layers no longer justify their limitations. He notes Shopify announced in September that "native is now the future of mobile at Shopify," citing AI agents, and rebuilt its Shop app in Swift and Kotlin in twelve weeks, with Coinbase apparently next. He keeps the Rails API as the security perimeter, treating the mobile apps as untrusted clients.

  40. Hacker News AI58

    Nvidia Crashed 9.3%. Is the AI Data-Center Boom Starting to Crack?

    On November 5, 2026, Nvidia closed down 9.26% at $181.60, AMD fell 17.26% and the Nasdaq Composite dropped 5.52%, with Reuters estimating roughly $2.5 trillion in Nasdaq market value erased; the Philadelphia Semiconductor Index fell about 12.25%. The article attributes the sell-off to three converging risks: data-center policy uncertainty after the November 3 midterms, questions over whether AI capital expenditure can generate timely cash returns, and broader earnings and rate pressure. It argues the move reflects repricing of listed shares rather than a collapse in AI chip shipments, and points to operational megawatts, AI revenue versus capex, and equipment order conversion as indicators to watch.