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

Wednesday, 16 September 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 2026-09-16 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 3

    Introducing System One Models & Jev

    TypeSafe AI launches Jev, a System One model that makes structured decisions 40x to 200x faster than existing LLMs. Includes a pricing table showing $0.042 input costs and free outputs, plus workflow evals comparing Jev against GPT-5.6 Terra. cited

    Reported byThe Rundown AI, TLDR AI, AI Breakfast

    Where it was shared
    typesafe.ai announcement · 10 min read First shared 2026-09-16
  2. Shared by 2

    Introducing Odyssey-3: A General-Purpose Physical Intelligence

    “Odyssey introduced Odyssey 3, a world model that can control robot arms, humanoids, self-driving cars, drones, and games, set to release in the coming weeks.” via The Rundown AI

    Reported byThe Rundown AI, TLDR AI

    How 1 source put it
    • The Rundown AI2026-09-16issue ↗ Odyssey introduced Odyssey 3, a world model that can control robot arms, humanoids, self-driving cars, drones, and games, set to release in the coming weeks.
    • TLDR AI2026-09-16issue ↗
    odyssey.systems First shared 2026-09-16
  3. Shared by 2

    Meta One is a subscription service that offers enhanced AI and self-expression features across Instagram, Facebook, and…

    “Meta rolled out Meta One, a new paid tier across its apps running from $2.99 to $499 a month for extra Meta AI usage and creator tools.” via The Rundown AI

    Reported byThe Rundown AI, TLDR AI

    How 1 source put it
    • The Rundown AI2026-09-16issue ↗ Meta rolled out Meta One, a new paid tier across its apps running from $2.99 to $499 a month for extra Meta AI usage and creator tools.
    • TLDR AI2026-09-16issue ↗
    about.fb.com First shared 2026-09-16
  4. Shared by 2

    Breaking the 1.58-bit Barrier for Ternary LLMs

    “Intel's BITCOS packs ternary weights at 1.485 bits each, under the 1.625 bits of five-trit packing, for up to 1.18x CPU and 1.27x GPU inference speedups (arXiv)” via AI Breakfast

    Reported byAI Breakfast

    Discussed onHacker News 245 pts · 41 comments

    arxiv.org First shared 2026-09-16
  5. Shared by 2

    404 Media reported OpenAI contractors read and rate real ChatGPT conversations under "Project Lily," with OpenAI acknowledging sensitive details can still get through redaction

    “A 404 Media report detailed OpenAI’s ‘Project Lily’, finding that hundreds of human contractors read and rate real ChatGPT chats, often without the user’s knowledge.” via The Rundown AI

    Reported byThe Rundown AI, AI Breakfast

    How 1 source put it
    • The Rundown AI2026-09-16issue ↗ A 404 Media report detailed OpenAI’s ‘Project Lily’, finding that hundreds of human contractors read and rate real ChatGPT chats, often without the user’s knowledge.
    • AI Breakfast2026-09-16issue ↗
    404media.co First shared 2026-09-16
  6. Shared by 2

    Learning to Solve Hard Problems in RL for LLMs by Never Giving Up

    “Standard RL results in disproportionately poor performance on the hardest problems. Simple scalar values may not be sufficient for accurate evaluations of large language models. This paper proposes a solution that reduces the compute spent on easy problems, which reallocates…” via TLDR AI

    Reported byTLDR AI

    Discussed onHacker News 119 pts · 9 comments

    mnoukhov.github.io First shared 2026-09-15
  7. Shared by 2

    Gemini 3.8 Live and 3.8 Live Extended Thinking

    “Google debuted Gemini 3.8 Live, voice models that can continue speaking while they think, with the Extended Thinking variant topping AA's speech-to-speech quality ranking.” via The Rundown AI

    Reported byThe Rundown AI

    Discussed onHacker News 493 pts · 328 comments

    blog.google First shared 2026-09-15
  8. Shared by 3

    Perplexity Portable Computer Is Now Available on Windows, Powered by NVIDIA RTX

    “Perplexity Portable Computer reached Windows PCs with supported NVIDIA graphics cards carrying at least 24GB of memory. It can work locally and asks permission before sending data to cloud models for more demanding reasoning. Users get a clearer boundary around which information leaves their computer. Announcement.” via Turing Post

    Reported byTuring Post, TLDR AI, AI Breakfast

    How 1 source put it
    • Turing Post2026-09-14issue ↗ Perplexity Portable Computer reached Windows PCs with supported NVIDIA graphics cards carrying at least 24GB of memory. It can work locally and asks permission before sending data to cloud models for more demanding reasoning. Users get a clearer boundary around which information leaves their computer. Announcement.
    • TLDR AI2026-09-15issue ↗
    • AI Breakfast2026-09-16issue ↗
    blogs.nvidia.com First shared 2026-09-14
  9. Shared by 1

    Nature Is Our Learning Environment

    “Periodic Neon outperforms GPT-6 Astra and Claude Fable 5.1 at lower cost on a highly challenging evaluation for scientific analysis. Neon is now deployed in labs to analyze experiments for better superconductors and magnets. The model establishes a Pareto-optimal cost-performance frontier through…” via TLDR AI

    Reported byTLDR AI

    periodic.com First shared 2026-09-16
  10. Shared by 1

    Gemini 3.8 Live and 3.5 Transcribe

    “Google released Gemini 3.8 Live and Gemini 3.5 Transcribe for developing real-time voice applications. These tools offer enhanced features for voice recognition, transcription, and live interaction capabilities. They aim to improve accuracy and user experiences in voice-driven applications.” via TLDR AI

    Reported byTLDR AI

    blog.google First shared 2026-09-16
  11. Shared by 1

    I made my website charge AI agents a penny per page. Then I watched Claude pay it

    “This developer has been testing an approach to charging AI agents for using their content. Their page sets a price through the x402 protocol before an agent reads it. If the agent wants access, it pays. While the developer has yet to receive any real payments,…” via TLDR AI

    Reported byTLDR AI

    suganthan.com First shared 2026-09-16
  12. Shared by 1

    Provides an open-source humanoid arm for teleoperation, imitation learning, and physical AI research.

    Reported byTLDR AI

    github.com First shared 2026-09-16
  13. Shared by 1

    Dream-RSI: Recursive Self-Improvement through Evolving Worlds

    “Progress in recursive self-improvement hinges on effective exploration. Accumulated discovery history can serve as a replay simulator over a realized search space. Dream-RSI closes a self-improvement loop at the exploration layer. It continuously collects discovery histories…” via TLDR AI

    Reported byTLDR AI

    paperswithcode.co First shared 2026-09-16
  14. Shared by 1

    Early Anthropic hire, former METR COO have found a way to rein in rogue AI agents

    “The Artificial Intelligence Underwriting Company (AIUC) hopes to bring AI safety to enterprises and companies building AI models and agents. It has built a third-party audit and certification layer for AI agents designed to give enterprises an independent…” via TLDR AI

    Reported byTLDR AI

    techcrunch.com First shared 2026-09-16
  15. Shared by 1

    See how Genie pushes the frontier for data agents →

    “Specialized knowledge search. Parallel thinking. Multiple LLMs. See how Databricks improved Genie's accuracy from 32% to over 90% on an internal benchmarkof real-world data analysis tasks. See how Genie pushes the frontier for data agents →” via TLDR AI

    Reported byTLDR AI

    databricks.com First shared 2026-09-16
  16. Shared by 1

    Programming is dead. Engineering is thriving

    “G5 Labs, an MIT CSAIL spinout, raised $14 million to develop a platform where natural language, structured as intent graphs, serves as the source code.” via TLDR AI

    Reported byTLDR AI

    g5labs.ai First shared 2026-09-16
  17. Shared by 1

    Noam Brown (@polynoamial) gave a very interesting interview on The Information on OpenAI's priorities and what comes next

    “OpenAI's clear priority is recursive self-improvement, and AI could surpass even Brown's own research intuition within one or two model releases from now.” via TLDR AI

    Reported byTLDR AI

    threadreaderapp.com First shared 2026-09-16
  18. Shared by 1

    Your Agent Aced the Task. Will It Do It Again?

    “ALTK-Evolve introduces Consistency Guidelines to address the inconsistency gap, where a model's task success rate drops significantly across repeated runs.” via TLDR AI

    Reported byTLDR AI

    huggingface.co First shared 2026-09-16
  19. Shared by 1

    The AI graveyard: a running list of projects and startups that didn't make it

    “Relay, an AI workflow tool, shut down due to competition from larger companies integrating similar features.” via TLDR AI

    Reported byTLDR AI

    techcrunch.com First shared 2026-09-16
  20. Shared by 2

    Chinese researchers detail 'the last AI built by humans'

    “AI RESEARCH🏗️ Chinese researchers detail 'the last AI built by humans'Image source: arXivThe Rundown: 30+ Chinese AI researchers, including those from ByteDance, Tsinghua, and Shanghai AI Lab, published “The Last AI Built by Humans,” a roadmap detailing five levels of recursive self-improvement, ending with AI that builds its successors.The details: Level 1…” via The Rundown AI

    Reported byTuring Post, The Rundown AI

    How 1 source put it
    • Turing Post2026-09-14issue ↗
    • The Rundown AI2026-09-16issue ↗ AI RESEARCH🏗️ Chinese researchers detail 'the last AI built by humans'Image source: arXivThe Rundown: 30+ Chinese AI researchers, including those from ByteDance, Tsinghua, and Shanghai AI Lab, published “The Last AI Built by Humans,” a roadmap detailing five levels of recursive self-improvement, ending with AI that builds its successors.The details: Level 1…
    arxiv.org First shared 2026-09-14

The week to that day two or more sources

  1. Shared by 8

    Dario Amodei

    Anthropic CEO Dario Amodei calls for slowing AI capability advancement to let safety measures catch up. Details a three-step plan including embedded third-party evaluators, democratic coordination on safety standards, and global government coordination. cited

    Reported byExponential View, TheSequence, The Rundown AI, Superhuman AI, TLDR AI, AI Breakfast, Turing Post, Forward Future

    Where it was shared
    darioamodei.com announcement · 17 min read First shared 2026-09-13
  2. Shared by 3

    Introducing System One Models & Jev

    TypeSafe AI launches Jev, a System One model that makes structured decisions 40x to 200x faster than existing LLMs. Includes a pricing table showing $0.042 input costs and free outputs, plus workflow evals comparing Jev against GPT-5.6 Terra. cited

    Reported byThe Rundown AI, TLDR AI, AI Breakfast

    Where it was shared
    typesafe.ai announcement · 10 min read First shared 2026-09-16
  3. Shared by 6

    Countering misuse of AI: September 2026 / Anthropic

    “Anthropic's Threat Intelligence team has identified and disrupted several operations in the past several months where threat actors tried to use Claude for malicious activity. This report shares case studies from those operations and describes how malicious use of Claude has evolved since…” via TLDR AI

    Reported byThe Rundown AI, Superhuman AI, TLDR AI, AI Breakfast, Interconnects, Ben's Bites

    How 1 source put it
    anthropic.com 156 min read First shared 2026-09-11
  4. Shared by 5

    OpenAI Launches the Agents API

    “OpenAI introduced the Agents API in public beta, giving developers access to the managed agent harness and infrastructure behind Codex. It handles context, tools, subagents, persistent execution, files, and code environments for agents that can run for extended periods.” via TLDR AI

    Reported byTLDR AI, AI Breakfast, TheSequence, Turing Post, Ben's Bites

    How 1 source put it
    openai.com First shared 2026-09-11
  5. Shared by 2

    Introducing Odyssey-3: A General-Purpose Physical Intelligence

    “Odyssey introduced Odyssey 3, a world model that can control robot arms, humanoids, self-driving cars, drones, and games, set to release in the coming weeks.” via The Rundown AI

    Reported byThe Rundown AI, TLDR AI

    How 1 source put it
    • The Rundown AI2026-09-16issue ↗ Odyssey introduced Odyssey 3, a world model that can control robot arms, humanoids, self-driving cars, drones, and games, set to release in the coming weeks.
    • TLDR AI2026-09-16issue ↗
    odyssey.systems First shared 2026-09-16
  6. Shared by 2

    Meta One is a subscription service that offers enhanced AI and self-expression features across Instagram, Facebook, and…

    “Meta rolled out Meta One, a new paid tier across its apps running from $2.99 to $499 a month for extra Meta AI usage and creator tools.” via The Rundown AI

    Reported byThe Rundown AI, TLDR AI

    How 1 source put it
    • The Rundown AI2026-09-16issue ↗ Meta rolled out Meta One, a new paid tier across its apps running from $2.99 to $499 a month for extra Meta AI usage and creator tools.
    • TLDR AI2026-09-16issue ↗
    about.fb.com First shared 2026-09-16
  7. Shared by 4

    David Sacks: OpenAI and Anthropic Don't Need Regulations to Pace Frontier Models

    “David Sacks told OpenAI and Anthropic to slow down unilaterally and "stop pretending you need anyone else's permission" (X)” via AI Breakfast

    Reported byAI Breakfast, The Rundown AI, Ben's Bites

    Discussed onHacker News 334 pts · 262 comments

    How 1 source put it
    x.com X post First shared 2026-09-13
  8. Shared by 3

    Perplexity Portable Computer Is Now Available on Windows, Powered by NVIDIA RTX

    “Perplexity Portable Computer reached Windows PCs with supported NVIDIA graphics cards carrying at least 24GB of memory. It can work locally and asks permission before sending data to cloud models for more demanding reasoning. Users get a clearer boundary around which information leaves their computer. Announcement.” via Turing Post

    Reported byTuring Post, TLDR AI, AI Breakfast

    How 1 source put it
    • Turing Post2026-09-14issue ↗ Perplexity Portable Computer reached Windows PCs with supported NVIDIA graphics cards carrying at least 24GB of memory. It can work locally and asks permission before sending data to cloud models for more demanding reasoning. Users get a clearer boundary around which information leaves their computer. Announcement.
    • TLDR AI2026-09-15issue ↗
    • AI Breakfast2026-09-16issue ↗
    blogs.nvidia.com First shared 2026-09-14
  9. Shared by 2

    Gemini 3.8 Live and 3.8 Live Extended Thinking

    “Google debuted Gemini 3.8 Live, voice models that can continue speaking while they think, with the Extended Thinking variant topping AA's speech-to-speech quality ranking.” via The Rundown AI

    Reported byThe Rundown AI

    Discussed onHacker News 493 pts · 328 comments

    blog.google First shared 2026-09-15
  10. Shared by 4

    Introducing SWE-2: Pushing the Pareto Frontier

    “Cognition rolled out SWE-2 inside Devin, a Kimi K3-based coding model that claims to match Fable 5.1 and GPT-6 Astra on certain coding benchmarks while costing 64% less.” via The Rundown AI

    Reported byThe Rundown AI, TLDR AI, AI Breakfast, Ben's Bites

    How 1 source put it
    cognition.com First shared 2026-09-11