Model Follower.

Up to date

Rebuilt
2026-10-25 06:54 UTC
Release radar
117 on watch
Hub snapshot
2026-10-25
Re-checked
daily · every number sourced and dated

Links / Past days

Saturday, 17 October 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 2026-10-17 23:59 UTC Status sealed Rebuilt from issues on file

Reconstructed after the day from the issues on file (dates as published). An issue counts on the date it carries, even if it was collected later.

That day 36-hour window

  1. Shared by 6

    Introducing Muse Spark Open, The Most Capable US Open‑Weight Model to Date

    Meta releases Muse Spark Open, an 816B-parameter multimodal model designed for coding and agentic work. Details how two Apple M5 Ultra Mac Studios run the model locally via MLX and lists hosted inference partners. cited

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

    Where it was shared
    research.meta.ai announcement · 3 min read First shared 2026-10-16
  2. Shared by 3

    Inside Microsoft's Frontier Pre-Training Stack

    “Microsoft detailed the training of MAI-Base-2, a 2.8T-parameter mixture-of-experts model with 80B active parameters that it said surpassed every open-weight model in its held-out pre-training evaluations. The deep dive covers its data strategy, compute efficiency, full-stack infrastructure, and…” via TLDR AI

    Reported byTLDR AI, AI Breakfast, Interconnects

    How 1 source put it
    • TLDR AI2026-10-16issue ↗ Microsoft detailed the training of MAI-Base-2, a 2.8T-parameter mixture-of-experts model with 80B active parameters that it said surpassed every open-weight model in its held-out pre-training evaluations. The deep dive covers its data strategy, compute efficiency, full-stack infrastructure, and…
    • AI Breakfast2026-10-16issue ↗
    • Interconnects2026-10-17issue ↗
    microsoft.ai First shared 2026-10-16
  3. Shared by 12

    Introducing Mistral Large 4

    Mistral launches a public preview of its 1 trillion-parameter, natively multimodal model Le Chonk. Includes cybersecurity benchmarks where it scores 82% on vulnerability patching and beats GPT-6-Astra on legal tasks. cited

    Reported bySimon Willison, The Neuron, The Rundown AI, TLDR AI, AI Breakfast, Last Week in AI, TheSequence, Import AI, Turing Post, The Batch, Interconnects

    Discussed onHacker News 522 pts · 5 comments

    Where it was shared
    mistral.ai announcement · 9 min read First shared 2026-10-06
  4. Shared by 1

    Inline rendering with document​.currentScript

    “Lean Rada describes a neat trick that takes advantage of the document.currentScript property. You can use that to drop in a tiny inline <script> element that replaces itself when the page loads: The year is <script> document.currentScript.replaceWith( new Date().getFullYear())</script> If you define a shortcut…” via Simon Willison

    Reported bySimon Willison

    leanrada.com First shared 2026-10-17
  5. Shared by 1

    You can use that to drop in a tiny inline <script> element that replaces itself when the page loads: The year is…

    “…ean Rada describes a neat trick that takes advantage of the document.currentScript property. You can use that to drop in a tiny inline <script> element that replaces itself when the page loads: The year is <script> document.currentScript.replaceWith( new Date().getFullYear())</script> If you define a shortcut function like this:…” via Simon Willison

    Reported bySimon Willison

    developer.mozilla.org First shared 2026-10-17
  6. Shared by 10

    Introducing Beam: Reflection’s 501B open-weight model

    Reflection AI launches Beam, a 501B sparse MoE model with 23B active parameters trained on 23.8T tokens. Details how 10.5K NVIDIA GB300 GPUs generated over 100 million rollouts to achieve 3–4× less inference compute than Qwen 3.8-Max. cited

    Reported byTuring Post, Latent Space / AINews, The Rundown AI, The Neuron, Superhuman AI, TLDR AI, TheSequence, Import AI, Interconnects

    Discussed onHacker News 556 pts · 173 comments

    Where it was shared
    reflection.ai announcement · 12 min read First shared 2026-10-05
  7. Shared by 8

    Muse Wonder: A Frontier Model For Personal & Professional AI Agents

    Meta releases Muse Wonder and Muse Wonder Full, its most capable frontier models for personal and professional agents. Details Active Memory and Active Focus features, plus benchmark results comparing performance on law, finance, and coding tasks. cited

    Reported byAI Breakfast, Simon Willison, The Rundown AI, Latent Space / AINews, The Neuron, TLDR AI, Ben's Bites, Interconnects

    Where it was shared
    research.meta.ai announcement · 4 min read First shared 2026-10-12
  8. Shared by 1

    For some background, from the Muse Spark Open model card on Hugging Face

    Reported byInterconnects

    huggingface.co First shared 2026-10-17
  9. Shared by 1

    When Meta released Muse Glimmer, the 30B agentic model, Alex Wang said that they were preparing to release the Muse…

    “In reality, this was much more likely a safety tune of Muse Spark 1.2, which would only be a 1 point drop on Artificial Analysis to shore up safety behaviors. When Meta released Muse Glimmer, the 30B agentic model, Alex Wang said that they were preparing to release the Muse Spark 1.2 model as open weights, which fits with the above.” via Interconnects

    Reported byInterconnects

    AboutMuse-Glimmer-30B

    huggingface.co First shared 2026-10-17
  10. Shared by 1

    If anything will hold Muse Spark Open back, it’ll be the specific license they designed for it.

    Reported byInterconnects

    huggingface.co First shared 2026-10-17
  11. Shared by 1

    Meta is the only American company with frontier open and closed models.

    “… major releases are coming just over a year out from Meta’s massive AI redirection, characterized by CEO Mark Zuckerberg’s Personal Superintelligence post, which got a lot of criticism at the time. Meta is the only American company with frontier open and closed models. Alibaba’s Qwen has followed a similar strategy, but with a far more open-leaning…” via Interconnects

    Reported byInterconnects

    meta.com First shared 2026-10-17
  12. Shared by 1

    I built a programming language that LLMs can't use

    Discussed onHacker News 282 pts · 218 comments

    blog.picheta.me First shared 2026-10-17
  13. Shared by 1

    Release: web-edit 0.1a2 web-edit (previously) is my tiny web app for browsing and editing files in a folder.

    Reported bySimon Willison

    github.com First shared 2026-10-17
  14. Shared by 2

    How the Claude Code team does code review / claude.dev Blog

    Claude Code merged PRs doubled from June to September while reverts fell over 40 percent. Includes a risk tier table mapping low, medium, and high changes to zero, one, or two human approvals. cited

    Reported bySimon Willison

    Discussed onHacker News 487 pts · 480 comments

    claude.dev guide · 13 min read First shared 2026-10-16
  15. Shared by 1

    We checked 12 subreddits, 544 Twitters and no further Discords.

    “AI News for 9/29/2026-9/30/2026. We checked 12 subreddits, 544 Twitters and no further Discords. AINews’ website lets you search all past issues. As a reminder, AINews is now a section of Latent Space. You can opt in/out of email frequencies!” via Latent Space / AINews

    Reported byLatent Space / AINews

    x.com First shared 2026-10-01
  16. Shared by 1

    RuneScape Bench: MMOs for Agents [video]

    Discussed onHacker News 249 pts · 82 comments

    youtube.com Video First shared 2026-10-17
  17. Shared by 1

    When Meta released Muse Glimmer, the 30B agentic model, Alex Wang said that they were preparing to release the Muse…

    “In reality, this was much more likely a safety tune of Muse Spark 1.2, which would only be a 1 point drop on Artificial Analysis to shore up safety behaviors. When Meta released Muse Glimmer, the 30B agentic model, Alex Wang said that they were preparing to release the Muse Spark 1.2 model as open weights, which fits with the above.” via Interconnects

    Reported byInterconnects

    AboutMuse-Glimmer-30B

    x.com X post First shared 2026-10-17
  18. Shared by 1

    UCLA scientists have unveiled PhoneBot, a 19-inch walking robot that uses an Android phone as its torso, tapping its…

    “…gns we’ve seen in a while, hands down. UCLA scientists have unveiled PhoneBot, a 19-inch walking robot that uses an Android phone as its torso, tapping its camera, motion sensors, and processor to see, balance, and move. Beyond the phone, parts cost about $400, presenting enthusiasts with an affordable entry point into robotics. You can see the unique (and…” via Superhuman AI

    Reported bySuperhuman AI

    techxplore.com First shared 2026-10-17
  19. Shared by 1

    Zuck and Alexandr Wang on The Cheeky Pint

    “…uddled in their offerings than Meta today. If you listen to Zuck and Alexandr Wang on The Cheeky Pint, they clearly had a very specific vision of a high-ownership, high-compute per researcher lab that I would expect to pay off. Many labs die due to layers of bureaucracy and misaligned incentives. Meanwhile Alex Wang has over 200 direct reports — i.e. it’s a…” via Interconnects

    Reported byInterconnects

    youtube.com Video First shared 2026-10-17
  20. Shared by 1

    Advertise to 700K readers of The Neuron here!

    “…l mistake can spiral into a much bigger mess. Let's dig in. Advertise in The Neuron here The AI agent that lied to a person to get its wayPicture a massive, shared library of free code that programmers all over the world use and contribute to. If you want to change or add something, you submit your proposed edit, and a real human has to review and approve…” via The Neuron

    Reported byThe Neuron

    info.technologyadvice.com First shared 2026-09-26

The week to that day two or more sources

  1. Shared by 6

    Introducing Muse Spark Open, The Most Capable US Open‑Weight Model to Date

    Meta releases Muse Spark Open, an 816B-parameter multimodal model designed for coding and agentic work. Details how two Apple M5 Ultra Mac Studios run the model locally via MLX and lists hosted inference partners. cited

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

    Where it was shared
    research.meta.ai announcement · 3 min read First shared 2026-10-16
  2. Shared by 9

    Redesigning source control at Anthropic

    Anthropic replaced git with Scion, merging over 30,000 code changes daily after migrating their monorepo in five months. Details how Scion uses always-online agents, server-side commits, and sparse checkouts to avoid storing full repo copies locally. cited

    Reported byAlphaSignal, Simon Willison, Superhuman AI, AI Breakfast, The Neuron, Latent Space / AINews, Don't Worry About the Vase, Ben's Bites

    Discussed onHacker News 898 pts · 1,013 comments

    Where it was shared
    anthropic.com announcement · 16 min read First shared 2026-10-13
  3. Shared by 8

    Muse Wonder: A Frontier Model For Personal & Professional AI Agents

    Meta releases Muse Wonder and Muse Wonder Full, its most capable frontier models for personal and professional agents. Details Active Memory and Active Focus features, plus benchmark results comparing performance on law, finance, and coding tasks. cited

    Reported byAI Breakfast, Simon Willison, The Rundown AI, Latent Space / AINews, The Neuron, TLDR AI, Ben's Bites, Interconnects

    Where it was shared
    research.meta.ai announcement · 4 min read First shared 2026-10-12
  4. Shared by 12

    Introducing Mistral Large 4

    Mistral launches a public preview of its 1 trillion-parameter, natively multimodal model Le Chonk. Includes cybersecurity benchmarks where it scores 82% on vulnerability patching and beats GPT-6-Astra on legal tasks. cited

    Reported bySimon Willison, The Neuron, The Rundown AI, TLDR AI, AI Breakfast, Last Week in AI, TheSequence, Import AI, Turing Post, The Batch, Interconnects

    Discussed onHacker News 522 pts · 5 comments

    Where it was shared
    mistral.ai announcement · 9 min read First shared 2026-10-06
  5. Shared by 6

    Gemini 4 Argon: Built to solve your hardest problems

    Gemini 4 Argon launches globally today for Pro and Ultra subscribers via the Gemini app and API. Includes internal tests where Argon watched a human play cornhole, built simulation code, and sank seven consecutive shots on a robot. cited

    Reported byAlphaSignal, The Rundown AI, The Neuron, Superhuman AI, Forward Future, Ben's Bites

    Where it was shared
    blog.google announcement · 10 min read First shared 2026-10-13
  6. Shared by 7

    AI models may be capable of self-replicating without access to a datacenter

    “Anthropic released new research demonstrating that small AI models are getting closer to being able to copy themselves between everyday gadgets like smart ovens or TVs.” via The Rundown AI

    Reported byThe Rundown AI, The Neuron, TLDR AI, Latent Space / AINews, AI Breakfast, Don't Worry About the Vase

    Discussed onHacker News 118 pts · 116 comments

    How 1 source put it
    anthropic.com 22 min read First shared 2026-10-12
  7. Shared by 3

    Inside Microsoft's Frontier Pre-Training Stack

    “Microsoft detailed the training of MAI-Base-2, a 2.8T-parameter mixture-of-experts model with 80B active parameters that it said surpassed every open-weight model in its held-out pre-training evaluations. The deep dive covers its data strategy, compute efficiency, full-stack infrastructure, and…” via TLDR AI

    Reported byTLDR AI, AI Breakfast, Interconnects

    How 1 source put it
    • TLDR AI2026-10-16issue ↗ Microsoft detailed the training of MAI-Base-2, a 2.8T-parameter mixture-of-experts model with 80B active parameters that it said surpassed every open-weight model in its held-out pre-training evaluations. The deep dive covers its data strategy, compute efficiency, full-stack infrastructure, and…
    • AI Breakfast2026-10-16issue ↗
    • Interconnects2026-10-17issue ↗
    microsoft.ai First shared 2026-10-16
  8. Shared by 3

    Anthropic update on independent evaluators

    Anthropic embedded six METR employees with hardware and access matching its internal risk teams. Details how Anthropic can require redactions for confidential IP while METR publishes findings without approval. cited

    Reported byThe Rundown AI, The Neuron, TLDR AI

    Where it was shared
    anthropic.com announcement · 4 min read First shared 2026-10-16
  9. Shared by 6

    Strengthening our defenses against distillation attacks

    “(Activity: 335): The image is a mobile screenshot from Anthropic’s research post on distillation defenses, highlighting a passage alleging Zhipu’s GLM-5.3 distilled cyber capabilities from leading U.S. models, is unusually capable among open-weight cyber models, and has bypassable safeguards. The Reddit title…” via Latent Space / AINews

    Reported byLatent Space / AINews, The Rundown AI, Forward Future, AI Breakfast, Don't Worry About the Vase

    Discussed onr/LocalLLaMA 755 upvotes · 146 comments

    How 1 source put it
    anthropic.com 9 min read First shared 2026-10-13
  10. Shared by 3

    Amazon introduces 2 new AI products from Bee: Bee Frontier and Bee Eclipse

    Amazon launches Bee Frontier wearable and Bee Eclipse phone attachment for $74.99 and $99.99. Details how devices process audio in real time, delete recordings, and integrate with Alexa+ and Amazon.com. cited

    Reported byThe Rundown AI, The Neuron, Superhuman AI

    Where it was shared
    aboutamazon.com product page · 6 min read First shared 2026-10-16