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

Thursday, 5 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 2026-11-05 23:59 UTC Status sealed

That day 36-hour window

  1. Shared by 4

    ChatGPT Images 2.6

    “OpenAI released ChatGPT Images 2.6 with native GIF generation from text or photo prompts” via AlphaSignal

    Reported byAlphaSignal, The Neuron, Superhuman AI, Ben's Bites

    How 1 source put it
    openai.com First shared 2026-11-04
  2. Shared by 4

    Kimi K3.1 Pro

    “Moonshot AI released full open weights for Kimi K3.1 Pro on Hugging Face” via AlphaSignal

    Reported byAlphaSignal, Latent Space / AINews, The Neuron

    Discussed onHacker News 626 pts · 186 comments

    AboutKimi-K3.1-Pro

    How 1 source put it
    huggingface.co First shared 2026-11-04
  3. Shared by 2

    Get 58% off with code RUNDOWN*📞 Grok Bot - SpaceXSI’s AI agent, now able to make phone calls for users🚀 Hy4 - Tencent’s…

    “🔒 Incogni - Remove your personal data from the web, so scammers and identity thieves can’t access it. Get 58% off with code RUNDOWN*📞 Grok Bot - SpaceXSI’s AI agent, now able to make phone calls for users🚀 Hy4 - Tencent’s new flagship open-source model 🌝 Kimi K3.1 Pro - Moonshot AI’s powerful open-weight model” via The Rundown AI

    Reported byLatent Space / AINews, The Rundown AI

    How 1 source put it
    • Latent Space / AINews2026-11-05issue ↗
    • The Rundown AI2026-11-05issue ↗ 🔒 Incogni - Remove your personal data from the web, so scammers and identity thieves can’t access it. Get 58% off with code RUNDOWN*📞 Grok Bot - SpaceXSI’s AI agent, now able to make phone calls for users🚀 Hy4 - Tencent’s new flagship open-source model 🌝 Kimi K3.1 Pro - Moonshot AI’s powerful open-weight model
    hy.tencent.ai First shared 2026-11-05
  4. Shared by 2

    Conductor is joining Meta

    “Meta is acquiring Conductor (Blogpost) - the pioneers of the form factor that everyone copied this year.” via Latent Space / AINews

    Reported byLatent Space / AINews, The Rundown AI

    How 1 source put it
    conductor.build 1 min read First shared 2026-11-05
  5. Shared by 2

    Ripple-65B Preview, a Frontier Open-Weight Model

    Ricursive releases Ripple-65B Preview, a model achieving 80.52 on TerminalBench 2.1 for $32M. Details the thin-and-deep architecture, synthetic data pipelines, and laddering experiments that enabled this efficiency. cited

    Reported byThe Rundown AI

    Discussed onHacker News 334 pts · 123 comments

    ricursive.com announcement · 4 min read First shared 2026-11-04
  6. Shared by 2

    OpenAI to acquire Pierre Computer Company

    “OpenAI is acquiring Pierre, the startup behind the code review tools inside Codex, and plans to add its Code Storage service to the platform as well.” via The Rundown AI

    Reported byThe Rundown AI

    Discussed onHacker News 261 pts · 91 comments

    openai.com First shared 2026-11-04
  7. Shared by 2

    Ricursive claims it is the best US open-weight model on agentic coding, ahead of larger MoEs including Muse Spark…

    “Ripple-65B Preview (Ricursive): This is a 65B dense model pretrained from scratch. Ricursive claims it is the best US open-weight model on agentic coding, ahead of larger MoEs including Muse Spark Open-816B, Beam-501B, Inkling-975B and Nemotron 3 Ultra-550B (launch).” via Latent Space / AINews

    Reported byLatent Space / AINews, The Neuron

    How 1 source put it
    • Latent Space / AINews2026-11-05issue ↗ Ripple-65B Preview (Ricursive): This is a 65B dense model pretrained from scratch. Ricursive claims it is the best US open-weight model on agentic coding, ahead of larger MoEs including Muse Spark Open-816B, Beam-501B, Inkling-975B and Nemotron 3 Ultra-550B (launch).
    • The Neuron2026-11-05issue ↗
    x.com X post First shared 2026-11-05
  8. Shared by 3

    Revolut says self-hosting GLM on 56 Nvidia B200s cut its AI coding inference costs about 90%

    “British fintech Revolut slashes AI coding costs 10x by cutting out Claude and going self-hosted (6 minute read) Revolut runs GLM-5.3 on 56 Nvidia B200 GPUs.” via TLDR

    Reported byTLDR, AI Breakfast, Latent Space / AINews

    How 1 source put it
    theregister.com First shared 2026-11-04
  9. Shared by 1

    The Web Is More Than Enough, by Paul Ford

    “The web browser is very, very, very powerful software for applications, and the cost of accessing that power is going down seemingly every day. — Paul Ford, The Web Is More Than Enough Tags: browsers, paul-ford, web-development” via Simon Willison

    Reported bySimon Willison

    ftrain.com 5 min read First shared 2026-11-05
  10. Shared by 2

    Accelerating engineering for the physical world

    “Two employees used GPT-6 Astra to design a device in under a week; prototypes were built in under a month.” via Forward Future

    Reported byTLDR AI, Forward Future

    How 1 source put it
    openai.com First shared 2026-11-04
  11. Shared by 1

    thinking-machines-lab/tinker-cookbook: Post-training with Tinker

    “Trained via distillation from Inkling plus two weeks of RL; fine-tunable through Tinker.” via AlphaSignal

    Reported byAlphaSignal

    github.com 5 min read First shared 2026-10-21
  12. Shared by 1

    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

    AboutInkling

    thinkingmachines.ai announcement · 29 min read First shared 2026-11-05
  13. Shared by 1

    Available now on Tinker for inference and fine-tuning; Apache 2.0 weights planned after safety testing.

    Reported byAlphaSignal

    auth.thinkingmachines.ai First shared 2026-11-05
  14. Shared by 2

    The post notes a practical requirement of 32GB+ RAM to run these GGUFs locally.

    “Unsloth released Dynamic V3.1 GGUFs for Qwen3.8-Flash-Next with 15% higher accuracy than competing quants.” via AlphaSignal

    Reported byLatent Space / AINews, AlphaSignal

    AboutQwen3.8-Flash-Next

    How 1 source put it
    huggingface.co First shared 2026-10-15
  15. Shared by 2

    Qwen3.8-Flash-Next: How to Run Locally

    Qwen3.8-Flash-Next is a 125B parameter MoE model that runs locally on 32GB+ RAM without GPU VRAM. Includes Dynamic V3.1 quantization, engram mixing, and MTP support reaching 170 tokens/s on an RTX 6000 PRO GPU. cited

    Reported byAlphaSignal, Latent Space / AINews

    AboutQwen3.8-Flash-Next

    Where it was shared
    unsloth.ai guide · 7 min read First shared 2026-11-04
  16. Shared by 2

    NousResearch/hermes-agent: The agent that grows with you

    Nous Research released Hermes, an agent with a built-in learning loop that creates skills from experience and runs on a $5 VPS. Includes migration from OpenClaw, a closed learning loop with autonomous skill creation, and support for Telegram, Discord, and CLI interfaces. cited

    Reported byInterconnects, AlphaSignal

    Where it was shared
    github.com repo · 7 min read First shared 2026-10-25
  17. Shared by 1

    Release Hermes Agent v0.23.0 — The Hearth Release

    “Nous Research released Hermes Agent v0.23.0, The Hearth Release, consolidating roughly 1,650 PRs since v0.22.0.” via AlphaSignal

    Reported byAlphaSignal

    github.com 33 min read First shared 2026-11-05
  18. Shared by 1

    Bug Report: Privacy violation in User-Agent header · Issue #107757

    Anthropic admits Claude Code leaked user emails in User-Agent headers and is rolling out a fix. Details the accidental inclusion of account email in model context and a new built-in mod blocking web tool calls. cited

    Discussed onHacker News 894 pts · 297 comments

    github.com news report · 5 min read First shared 2026-11-05
  19. Shared by 9

    Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights

    GPT-6 Astra edits 0.016% of Qwen3.5-0.8B weights to raise MATH-500 from 7.9% to 48.1% without training. Includes demos where Astra identifies games from raw weights and writes MinAtar agents beating PPO 100M steps. cited

    Reported byTuring Post, Latent Space / AINews, The Neuron, The Rundown AI, Superhuman AI, Simon Willison, AlphaSignal, AI Breakfast

    Discussed onHacker News 2,951 pts · 896 comments

    Where it was shared
    rikiyat.github.io paper · 18 min read First shared 2026-11-02
  20. Shared by 1

    SimCSE: Simple Contrastive Learning of Sentence Embeddings

    “Vietnamese sentence embedding model combining SimCSE contrastive learning with PhoBERT backbone, 135M parameters.” via AlphaSignal

    Reported byAlphaSignal

    arxiv.org 2 min read First shared 2026-11-05

The week to that day two or more sources

  1. Shared by 9

    Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights

    GPT-6 Astra edits 0.016% of Qwen3.5-0.8B weights to raise MATH-500 from 7.9% to 48.1% without training. Includes demos where Astra identifies games from raw weights and writes MinAtar agents beating PPO 100M steps. cited

    Reported byTuring Post, Latent Space / AINews, The Neuron, The Rundown AI, Superhuman AI, Simon Willison, AlphaSignal, AI Breakfast

    Discussed onHacker News 2,951 pts · 896 comments

    Where it was shared
    rikiyat.github.io paper · 18 min read First shared 2026-11-02
  2. Shared by 4

    ChatGPT Images 2.6

    “OpenAI released ChatGPT Images 2.6 with native GIF generation from text or photo prompts” via AlphaSignal

    Reported byAlphaSignal, The Neuron, Superhuman AI, Ben's Bites

    How 1 source put it
    openai.com First shared 2026-11-04
  3. Shared by 5

    Siri AI is now open to all developers

    iOS 27.3 betas let developers integrate apps with Siri AI using App Intents and skills. Includes details on App Intents, skills for combining actions, and Spotlight semantic index access. cited

    Reported byThe Neuron, The Rundown AI, TLDR, TLDR AI, AI Breakfast

    Where it was shared
    developer.apple.com announcement · 3 min read First shared 2026-11-04
  4. Shared by 5

    Progress on provable inference for large language models

    Anthropic releases FlashProve, a system generating cryptographic proofs for a 400B-parameter model at 110 tokens/second. Includes executable code, Lean formalizations, and a protocol specification under the Apache 2.0 license. cited

    Reported byAlphaSignal, The Rundown AI, TLDR AI, AI Breakfast

    Discussed onHacker News 236 pts · 121 comments

    Where it was shared
    anthropic.com announcement · 20 min read First shared 2026-11-03
  5. Shared by 4

    Kimi K3.1 Pro

    “Moonshot AI released full open weights for Kimi K3.1 Pro on Hugging Face” via AlphaSignal

    Reported byAlphaSignal, Latent Space / AINews, The Neuron

    Discussed onHacker News 626 pts · 186 comments

    AboutKimi-K3.1-Pro

    How 1 source put it
    huggingface.co First shared 2026-11-04
  6. Shared by 4

    The 'Astra can read and write weights to improve Qwen' post is misleading

    Matthew Honnibal argues the Astra Qwen post is misleading because it conflates raw weight reading with Python-based editing. Details a table showing only the Qwen experiment used Python to run the model, while true raw weight reading occurred on smaller networks. cited

    Reported bySimon Willison, TLDR AI, Latent Space / AINews

    Discussed onHacker News 1,036 pts · 209 comments

    Where it was shared
    honnibal.dev analysis · 4 min read First shared 2026-11-03
  7. Shared by 6

    Anthropic warns that it's very close to RSI capability, aka self-building AI

    Anthropic's updated S-1 states it may imminently meet its definition of dramatically accelerated progress attributable to AI. Details the company's admission that it has already used AI to develop new models whose leaps in capability it didn't recognize at the time. cited

    Reported byForward Future, TLDR, TLDR AI, The Deep View, Mindstream

    Discussed onTechmeme

    Where it was shared
    techcrunch.com news report · 3 min read First shared 2026-11-02
  8. Shared by 5

    OpenAI grades its AI on everyday tasks

    “OpenAI unveiled LifeVal, a benchmark grading models on everyday personal-assistant tasks.” via AlphaSignal

    Reported byAlphaSignal, The Rundown AI, The Deep View, TLDR AI, AI Breakfast

    How 1 source put it
    openai.com First shared 2026-11-02
  9. Shared by 6

    We’ve received our first response from the Project Suncatcher prototype satellite.

    Google received the first AI response from its Project Suncatcher satellite after uplinking a prompt to the onboard Gemma 4 model. Includes the haiku generated by the model as it passed Svalbard and plans to test thermal stress and radiation exposure next. cited

    Reported byThe Rundown AI, The Deep View, TLDR AI, AI Breakfast, FryAI, AlphaSignal

    Aboutgemma-4-26B-A4B-it-assistant, gemma-4-31B-it-assistant

    Where it was shared
    blog.google news report · 1 min read First shared 2026-11-02
  10. Shared by 4

    OpenAI: ChatGPT Search Is About 20% of Google's, 250 Million People Speak With ChatGPT, and Chat Is Merging With Work

    “OpenAI said ChatGPT handles about one-fifth of Google's search-style queries, and weekly users rose from 1B to 1.2B in about 60 days.” via The Neuron

    Reported byThe Neuron, TLDR, TLDR AI, AI Breakfast

    How 1 source put it
    bigtechnology.com First shared 2026-11-04