Model Follower.

Up to date

Rebuilt
2026-11-15 07:14 UTC
Release radar
108 on watch
Hub snapshot
2026-11-15
Re-checked
daily · every number sourced and dated

Links / November 2026

Tuesday, 10 November 2026

The AI links the most newsletters and link sources had shared by the end of the day (UTC), each source counted once.

As of 2026-11-10 23:59 UTC Status sealed

That day 36-hour window

  1. Shared by 12

    Introducing Claude Fable 5.2 and Claude Mythos 5.2

    Anthropic released Claude Fable 5.2 and Mythos 5.2, which generate outputs up to 30% faster than the previous versions. Includes a benchmark table showing Fable 5.2 scores 49.1% on FrontierCode 2.0 Xhigh effort and details pricing of $8 per million input tokens. cited

    Reported byAlphaSignal, Simon Willison, Turing Post, Latent Space / AINews, The Neuron, The Rundown AI, Superhuman AI, The Deep View, TLDR AI, Ben's Bites, What's Up in AI

    Discussed onHacker News 991 pts · 8 comments

    Where it was shared
    anthropic.com product page · 18 min read First shared 2026-11-09
  2. Shared by 6

    Introducing Our Smallest, Most Efficient Model: Muse Ember

    Meta launches Muse Ember, a multimodal model costing $0.10 per million input tokens for agent subtasks. Details evaluation results showing a 37.8% score on AutomationBench-AAE2E business workflows and 40.9% on Terminal-Bench 4.0. cited

    Reported byAlphaSignal, Turing Post, Simon Willison, The Neuron, The Rundown AI, TLDR AI

    Where it was shared
    research.meta.ai announcement · 9 min read First shared 2026-11-09
  3. Shared by 4

    Our newest TPUs: 8tx and 8i Max, built for the agentic era

    Google announces TPU 8tx for confidential inference and TPU 8i Max for massive context windows. Details how 8tx keeps data protected in memory while 8i Max doubles compute and memory bandwidth per rack. cited

    Reported byTuring Post, The Rundown AI, Forward Future, TLDR AI

    Where it was shared
    blog.google announcement · 8 min read First shared 2026-11-10
  4. Shared by 3

    Gemini 4 Carbon is now available

    Gemini 4 Carbon is now available to all developers and enterprises at the same introductory price as Argon. Includes benchmark results showing #1 ranking on CursorBench 4.0 and 100% recall on internal scanning benchmarks. cited

    Reported byAlphaSignal, Simon Willison

    Discussed onHacker News 282 pts · 155 comments

    Where it was shared
    blog.google announcement · 8 min read First shared 2026-11-10
  5. Shared by 3

    Adobe and The Walt Disney Company Announce Multi-Year Strategic Alliance to Support Disney Innovation with New AI-Powered Tools

    Adobe and Disney form a multi-year alliance to deploy AI tools across theme parks, ESPN, and film studios. Details how Disney uses Firefly Foundry for sketch-to-image models and custom assets, plus StudioLAB and ESPN Edge collaborations. cited

    Reported byThe Neuron, The Rundown AI, The Deep View

    Where it was shared
    news.adobe.com news report · 4 min read First shared 2026-11-10
  6. Shared by 3

    Prompting Claude Fable 5.2

    Anthropic documents specific prompting patterns for Claude Fable 5.2, including effort calibration and code ownership instructions. Includes a system prompt sentence that caused the model to merge repeated code in about half of the tries during testing. cited

    Reported bySimon Willison, Superhuman AI, What's Up in AI

    Where it was shared
    platform.claude.com guide · 18 min read First shared 2026-11-10
  7. Shared by 3

    twist ending

    “A Wharton professor asked Fable 5.2 to create the most impressive thing possible. See what it created, and read his full review of Anthropic’s latest model.” via Superhuman AI

    Reported bySuperhuman AI, TLDR, What's Up in AI

    How 1 source put it
    oneusefulthing.org First shared 2026-11-10
  8. Shared by 5

    The components of agent memory

    Anthropic shares how Claude agents use idle time to reorganize memory and fix errors through a process called dreaming. Includes evaluations of CL-Bench 1.0 and Pokémon Red showing how specific models like Opus 5.5 close learning loops and avoid navigation traps. cited

    Reported byThe Rundown AI, TLDR AI, Turing Post, Ben's Bites

    Discussed onHacker News 189 pts · 104 comments

    Where it was shared
    anthropic.com analysis · 16 min read First shared 2026-11-07
  9. Shared by 2

    Anthropic joins AI’s natural gas frenzy, buys $3.4B worth of turbines

    Anthropic spends $3.4 billion on Caterpillar generators to produce up to 1.7 gigawatts of electricity. Details how Anthropic took over equipment orders from supplier OnePWR due to financing issues and grid connection queues. cited

    Reported byForward Future, TLDR AI

    Where it was shared
    techcrunch.com news report · 2 min read First shared 2026-11-10
  10. Shared by 2

    Update on October 2026 security incident

    Fireworks discovered impacts to account and billing information plus incomplete models for a limited set of accounts. Details that payment card numbers are not stored by Fireworks and were not impacted during this breach. cited

    Reported byThe Neuron, TLDR AI

    Where it was shared
    fireworks.ai news report · 3 min read First shared 2026-11-10
  11. Shared by 2

    ChatGPT Enterprise on AWS Marketplace

    “OpenAI put ChatGPT Enterprise in AWS Marketplace, so AWS customers can buy it through existing cloud billing and spending commitments.” via The Neuron

    Reported byThe Neuron, TLDR AI

    How 1 source put it
    • The Neuron2026-11-10issue ↗ OpenAI put ChatGPT Enterprise in AWS Marketplace, so AWS customers can buy it through existing cloud billing and spending commitments.
    • TLDR AI2026-11-10issue ↗
    openai.com First shared 2026-11-10
  12. Shared by 2

    Claude Whiteboards, Claude Sheets, and Claude Data

    Anthropic launches Claude Whiteboards and Sheets in beta while making Claude Data generally available. Details how teams can brainstorm on boards, edit shared spreadsheets with change tracking, and run certified data queries with live sharing. cited

    Reported byAlphaSignal

    Discussed onHacker News 187 pts · 142 comments

    claude.com announcement · 4 min read First shared 2026-11-10
  13. Shared by 2

    Read the full New York Times report

    “Nvidia’s top lawyer reportedly emailed agents in August 2025 as they investigated whether Singapore’s Megaspeed smuggled its chips to China, with the U.S. Commerce Department then halting the case and handing it to the Justice Department.” via The Rundown AI

    Reported byThe Rundown AI, Forward Future

    How 1 source put it
    • The Rundown AI2026-11-10issue ↗ Nvidia’s top lawyer reportedly emailed agents in August 2025 as they investigated whether Singapore’s Megaspeed smuggled its chips to China, with the U.S. Commerce Department then halting the case and handing it to the Justice Department.
    • Forward Future2026-11-10issue ↗
    nytimes.com First shared 2026-11-10
  14. Shared by 2

    OpenAI's Control Room Can Halt Research Model Output

    “OpenAI launched Control Room, a centralized kill-switch for every research model workload.” via AlphaSignal

    Reported byAlphaSignal, TLDR AI

    How 1 source put it
    openai.com First shared 2026-11-09
  15. Shared by 1

    Amplifying | Coding agent intelligence

    “…first seven times, yet showed up 55 times as an alternative. If the developer asks an agent to implement email and accepts the patch, how often does anyone get around to trying the second choice? The study breaks down the results by model and app. Our more recent comprehensive AEO study was a bit more competitive, with Postmark coming up top for most…” via Latent Space / AINews

    Reported byLatent Space / AINews

    amplifying.ai 32 min read First shared 2026-11-10
  16. Shared by 2

    GR00T 2: Scaling Robot Foundation Models for Generalization, Dexterity, and Adaptation

    GR00T 2 predicts future video and robot actions together, achieving 40% zero-shot success across 50 unseen tasks. Includes 95% success after five demonstrations, 88–96% contact-rich task success via RL, and 97/100 assembly completions with an agent. cited

    Reported byTuring Post, TLDR AI

    Where it was shared
    research.nvidia.com paper · 13 min read First shared 2026-11-10
  17. Shared by 1

    NaiveAI-Labs/Naive-N0.5-Flash: Naive-N0.5-Flash is an open-weight 309B MoE model with 15.5B active parameters, built for coding and AI R&D.

    NaiveAI releases an open-weight 309B MoE model with 15.5B active parameters and native 1M-token context. Includes AI-optimized inference up to 2,000 tokens/s, FP8 weights requiring 315 GB, and API pricing of $0.10/$0.40/$0.01 per million tokens. cited

    Reported byLatent Space / AINews

    github.com repo · 5 min read First shared 2026-11-10
  18. Shared by 1

    NaiveAI

    “Naive-N0.5-Flash is a new open-weights model built for AI coding tasks and tuned to generate as many tokens per second as possible. It was developed by a Chinese company called NaiveAI, with the aim of speeding up engineering and research workflows so that teams can experiment, iterate and solve complex problems faster.” via Latent Space / AINews

    Reported byLatent Space / AINews

    naive.ai 1 min read First shared 2026-11-10
  19. Shared by 1

    Xiaomi MiMo-V2.5

    Xiaomi releases MiMo-V2.5, a 310B Sparse MoE model supporting 1 million tokens of context. Includes benchmarks showing 62.3 on Claw-Eval and details the five-stage training pipeline from text pre-training to RL. cited

    Reported byLatent Space / AINews

    AboutMiMo-V2.5

    mimo.xiaomi.com announcement · 2 min read First shared 2026-11-10
  20. Shared by 2

    Claude Fable 5.2 launch offers

    Subscribers on eligible paid plans receive extra Fable usage and limit resets starting November 9, 2026. Details the specific resets for Max/Ultra plans based on subscription length and the one-week inclusion for Pro plans. cited

    Reported bySuperhuman AI, What's Up in AI

    Where it was shared
    support.claude.com announcement · 6 min read First shared 2026-11-10

The week to that day two or more sources

  1. Shared by 12

    Introducing Claude Fable 5.2 and Claude Mythos 5.2

    Anthropic released Claude Fable 5.2 and Mythos 5.2, which generate outputs up to 30% faster than the previous versions. Includes a benchmark table showing Fable 5.2 scores 49.1% on FrontierCode 2.0 Xhigh effort and details pricing of $8 per million input tokens. cited

    Reported byAlphaSignal, Simon Willison, Turing Post, Latent Space / AINews, The Neuron, The Rundown AI, Superhuman AI, The Deep View, TLDR AI, Ben's Bites, What's Up in AI

    Discussed onHacker News 991 pts · 8 comments

    Where it was shared
    anthropic.com product page · 18 min read First shared 2026-11-09
  2. Shared by 6

    Introducing Our Smallest, Most Efficient Model: Muse Ember

    Meta launches Muse Ember, a multimodal model costing $0.10 per million input tokens for agent subtasks. Details evaluation results showing a 37.8% score on AutomationBench-AAE2E business workflows and 40.9% on Terminal-Bench 4.0. cited

    Reported byAlphaSignal, Turing Post, Simon Willison, The Neuron, The Rundown AI, TLDR AI

    Where it was shared
    research.meta.ai announcement · 9 min read First shared 2026-11-09
  3. Shared by 10

    Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips

    Google makes Gemini Omni 1.2 Flash generally available at $0.07 per second for 720p video. Includes native 15-second generation, multi-turn conversational editing, and improved reference adherence for consistent characters and products. cited

    Reported byAlphaSignal, Latent Space / AINews, The Rundown AI, TLDR AI, Superhuman AI, The Neuron, AI Breakfast, Turing Post, FryAI

    Discussed onHacker News 369 pts · 253 comments

    Where it was shared
    blog.google announcement · 4 min read First shared 2026-11-05
  4. Shared by 4

    Our newest TPUs: 8tx and 8i Max, built for the agentic era

    Google announces TPU 8tx for confidential inference and TPU 8i Max for massive context windows. Details how 8tx keeps data protected in memory while 8i Max doubles compute and memory bandwidth per rack. cited

    Reported byTuring Post, The Rundown AI, Forward Future, TLDR AI

    Where it was shared
    blog.google announcement · 8 min read First shared 2026-11-10
  5. Shared by 5

    The components of agent memory

    Anthropic shares how Claude agents use idle time to reorganize memory and fix errors through a process called dreaming. Includes evaluations of CL-Bench 1.0 and Pokémon Red showing how specific models like Opus 5.5 close learning loops and avoid navigation traps. cited

    Reported byThe Rundown AI, TLDR AI, Turing Post, Ben's Bites

    Discussed onHacker News 189 pts · 104 comments

    Where it was shared
    anthropic.com analysis · 16 min read First shared 2026-11-07
  6. Shared by 7

    Introducing Inkling 2 Small

    Thinking Machines releases Inkling 2 Small, a 594B MoE model with 17B active parameters that supports image inputs. Includes benchmark tables showing it scores 31 on the Artificial Analysis Intelligence Index and 1527 ELO on AA-Briefcase. cited

    Reported byAlphaSignal, Latent Space / AINews, TLDR AI, The Rundown AI, The Deep View, Superhuman AI, Turing Post

    AboutInkling

    Where it was shared
    thinkingmachines.ai announcement · 29 min read First shared 2026-11-05
  7. Shared by 5

    A proof of the Aaronson-Ambainis conjecture

    Physical Superintelligence used an Anthropic model to prove the Aaronson-Ambainis conjecture, a 2008 quantum computing problem. Includes the full proof accepted by Lean, details on how the system improved the result twice, and Stanford researchers' comments. cited

    Reported byThe Neuron, The Rundown AI, The Deep View, TLDR AI

    Discussed onTechmeme

    Where it was shared
    psi.inc news report · 7 min read First shared 2026-11-06
  8. Shared by 8

    Introducing Personal Finance in Claude

    Anthropic launches Personal Finance in Claude for US Pro, Max, and Ultra subscribers to analyze up to two years of read-only transaction data. Includes examples of spending breakdowns, upcoming payment alerts, and APR calculations, while noting the tool cannot move or invest money. cited

    Reported byAlphaSignal, The Rundown AI, TLDR AI, Forward Future, Superhuman AI, The Deep View, The Neuron, FryAI

    Where it was shared
    claude.com announcement · 3 min read First shared 2026-11-05
  9. Shared by 3

    Gemini 4 Carbon is now available

    Gemini 4 Carbon is now available to all developers and enterprises at the same introductory price as Argon. Includes benchmark results showing #1 ranking on CursorBench 4.0 and 100% recall on internal scanning benchmarks. cited

    Reported byAlphaSignal, Simon Willison

    Discussed onHacker News 282 pts · 155 comments

    Where it was shared
    blog.google announcement · 8 min read First shared 2026-11-10
  10. Shared by 7

    OpenAI’s new model looks for software flaws

    “OpenAI just released GPT-6 Cyber, a version of GPT-6 Astra built to find security holes and write exploits: code that takes advantage of a specific bug. OpenAI rates it “Critical” for cybersecurity, the highest risk tier in its internal framework. Not exactly a subtle job description.” via The Neuron

    Reported byThe Neuron, The Rundown AI, TLDR AI, AI Breakfast, The Deep View, FryAI, Turing Post

    How 1 source put it
    openai.com First shared 2026-11-06