No. 152

AI Daybook

This issue 22 stories · 13 outlets1 repo5 papers

You're reading a past issue. See today's issue →

Today's briefing

Open source & papers radar

Repos and papers in this issue
Sunday special

Weekly radar

Nov 9 – Nov 15 · 6 repos · 1 model · 3 papers
  1. 1RepoAgentsFriday issue

    modern-web-guidance-src keeps coding agents updated on web best practices

    GitHub TrendingRepository: GoogleChrome/modern-web-guidance-srcStars: 2,722New stars that day: 799Language: TypeScript
  2. 2RepoAgentsSaturday issue

    open-code-review, an LLM-based code review agent at Alibaba scale

    GitHub TrendingRepository: alibaba/open-code-reviewStars: 50,510New stars that day: 206Language: Go
  3. 3RepoAgentsFriday issue

    Qwen Code, an open-source AI coding agent for the terminal

    GitHub TrendingRepository: QwenLM/qwen-codeStars: 29,738New stars that day: 151Language: TypeScript
  4. 4RepoOpen Source & ToolsMonday issue

    claude-quickstarts, quick start projects for building with the Claude API

    GitHub TrendingRepository: anthropics/claude-quickstartsStars: 19,355New stars that day: 28Language: Python
  5. 5RepoPhysical AIMonday issue

    MuJoCo, a general purpose physics simulator for multi-joint dynamics

    GitHub TrendingRepository: google-deepmind/mujocoStars: 16,337New stars that day: 26Language: C++
  6. 6RepoOpen Source & ToolsThursday issue

    OpenAI releases Codex Security CLI and SDK for vulnerability scanning and remediation

    GitHub TrendingRepository: openai/codex-securityStars: 11,633New stars that day: 20Language: TypeScript
  7. 7ModelLLMFriday issue

    ZONOS3-Mini, a text-to-speech model from Zyphra

    Hugging FaceModel: Zyphra/ZONOS3-MiniParameters: about 886 millionTask: text-to-speechLikes: 223Downloads (last 30 days): 8,543
  8. 8PaperResearchSaturday issue

    PhysLoop harnesses LLM physics awareness to improve video generation models

    Hugging FacearXiv 2611.16748Upvotes: 41
  9. 9PaperPhysical AISaturday issue

    Ropedia Xperience 2 provides 1,589 3D scans and robot learning data for physical AI scaling

    Hugging FacearXiv 2611.17060Upvotes: 25
  10. 10PaperResearchSaturday issue

    WorldDraft: evaluating video models as generative renderers for 4D worlds

    Hugging FacearXiv 2611.16217Upvotes: 14

Picked from 55 repos, models and papers in this week's seven issues (Monday to Sunday). Ranked by importance, then trending score, then engagement (stars, likes, upvotes); each format gets two slots first so no single format takes over.

From Seoul

Korea Desk

What Korea’s tech press is reporting, summarized in English.

2 stories in this issue come from Korean outlets. Originals are in Korean.

Agents

3 stories · 1 on the front pageAll stories in this section →

Physical AI

9 stories · 5 on the front page · 1 at the Korea DeskAll stories in this section →

Open Source & Tools

3 stories · 1 on the front pageAll stories in this section →

Research

2 stories · 1 on the front pageAll stories in this section →

Industry & Policy

4 stories · 2 on the front page · 1 at the Korea DeskAll stories in this section →

Stories in this issue

22 stories
Nov 15
Industry & Policy

Anthropic CEO admits 'huge uncertainty' on costs days before record IPO

Days before Anthropic's planned initial public offering, CEO Dario Amodei called buying compute the company's "deepest business dilemma." Anthropic's commitments now total nearly $901 billion, while it remains unclear whether future revenue will justify those costs. The IPO is largely intended as a financial buffer against that uncertainty.

The Decoder
Nov 14
Physical AI

OpenAI reassigns robotics researchers as general-purpose models control robots directly

OpenAI is dissolving its standalone robotics team, redistributing researchers to other teams. General-purpose models like GPT-6 Astra can apparently control robots directly, making a dedicated unit unnecessary, the company said.

The Decoder
Nov 14
Physical AI

Runway's open-source Praxis-1 uses video AI to control robots

Runway released Praxis-1, a robotics model with open weights and source code under the Apache 2.0 license. The model applies knowledge from video training to robot control and leads the World Action Model category on Nvidia's RoboLab-120 benchmark, though it cannot yet plan complex action sequences independently.

The Decoder
Nov 14
Physical AI

Zipline raises $1.5 billion for nationwide drone delivery network expansion

Zipline announced a $1.5 billion funding round at a $21.5 billion post-money valuation, led by Paradigm, WCM Investment Management, and Layer Global. The capital will expand the company's U.S. manufacturing facilities.

Robotics & Automation News
Nov 14
Industry & Policy

Jensen Huang's departure blasts hole in AI startup funding, VCs warn

Nvidia poured tens of billions into AI startups under Huang. Without him at the helm, founders and venture capitalists are bracing for smaller checks and lower valuations.

TechCrunch
Nov 13
Physical AIPaper

SALA enables dual-axis scaling of embodied foundation models through semantic latent action alignment

The authors propose SALA to address the challenge of learning latent actions in embodied foundation models. Using language descriptions of unpaired human and robot video clips, SALA estimates semantic similarities and guides human latent actions toward semantically related robot latents. The resulting shared latent action space enables scaling through increasing human video pre-training and task-specific fine-tuning, improving downstream robot task performance.

Takara TLDR
Nov 15
Open Source & ToolsRepo

Univer: open-source AI office SDK for editing, collaboration, and file processing

Univer Office SDK is a extensible open-source platform that brings editing, calculation, collaboration, and file processing capabilities to products for both people and AI agents.

GitHub Trending +2 moreTrending 63
Nov 5
ResearchPaper

Self-improving embodied agents overfit verification criteria, harming task transfer

The authors study whether coding agents exposed to benchmark criteria generate code that better performs the instructed task or overfits to verification criteria. They find that criteria exposure raises initial success rates but collapses when criteria change slightly, and skills transfer poorly to new tasks. They propose instruction-faithful criteria as a solution.

Hugging FaceTrending 11
Oct 16
Physical AIPaper

GPT-6-Astra achieves strong performance across robotic benchmarks but faces deployment gaps

GPT-6-Astra demonstrates unprecedented performance across 12 robotic benchmarks spanning manipulation, humanoids, and mobile platforms through in-context learning. Close-loop coordination between high-level reasoning and low-level control is effective but costly. Real-world deployment reveals substantial gaps in collecting robot demonstrations and handling physical constraints.

arXivTrending 6
Oct 16
ResearchPaper

WEAVE transforms multi-PR code histories into dependency-structured tasks for agent training

The authors introduce WEAVE, a framework that structures verified pull request histories into executable, dependency-aware tasks for code agents. They develop Graph Credit, a metric that measures dependency-consistent partial progress and supports dense rewards for reinforcement learning on long-horizon code changes.

arXivTrending 4
Oct 15
Physical AIPaper

CarryOn augments vision-language-action models with personalization and safety assurance

CarryOn is a dual-loop harness that augments a frozen vision-language-action policy. Its planning loop maintains persistent user and scene context for personalized task execution, while its execution loop actively supervises rollouts and shields against hazardous actions. On a real robot, it reduced safety violations from 52.6% to 6.1% and increased recovery success from 13% to 80%.

arXivTrending 4
Nov 15
LLM

Anthropic: outdated prompting habits slow down Claude Opus 5.5

Anthropic made Claude Opus 5.5 the default model on all paid plans and published a guide on improved prompting. The main recommendation is to drop outdated "think hard" instructions and start with medium effort instead.

The Decoder
Nov 14
Agents

Vercel releases UI SDK with 50+ UI behaviors for AI coding agents

Vercel Labs released Vercel UI SDK, an Apache-2.0 beta library containing over 50 UI behaviors. Coding agents can compose them into accessible components for React, Vue, Svelte, Solid, and Web Components, with a bridge for direct UI control.

MarkTechPost
Nov 14
Physical AI

Wing invests $125 million in US drone manufacturing for nationwide expansion

Wing, the drone delivery company owned by Alphabet, is investing up to $125 million in US manufacturing over the next two years. The investment includes a new 15,000-square-foot manufacturing facility in Fort Worth, Texas, to support nationwide autonomous delivery expansion.

Robotics & Automation News
Nov 14
Physical AI

Neura Robotics partners with E.ON to deploy humanoid robots in energy infrastructure

Neura Robotics, E.ON, and German energy company enviaM announced a partnership to develop physical AI applications for electricity networks. Initial real-world trials are planned for 2027 and will focus on using intelligent robots to inspect and monitor energy infrastructure and support workers in physically demanding or hazardous tasks.

Robotics & Automation News
Nov 14
Industry & Policy

Amazon plans World Product Graph to index online shopping data globally

Amazon is building a database of products sold anywhere on the internet, including items it does not sell, and will make the data available through its own site. The project aims to direct shoppers away from Reddit and Google toward Amazon for product research, and to influence AI agents to rely on Amazon's product index.

The Verge AI
Nov 14
Agents

Anthropic upgrades Claude Code Projects to run parallel AI workers

Anthropic upgraded Claude Code Projects to enable parallel AI workers.

Hacker NewsTrending 53
Nov 15
Open Source & Tools

Cmpnd releases dspy-wsp for weight-space programming in DSPy

Cmpnd released dspy-wsp, a prototype DSPy optimizer for weight-space programming. A frontier LLM reads a small model's weights, edits them in a sandbox, and tests each change, with no gradients required and at a cost of about $5 to $10 per run.

MarkTechPost
Nov 14
Open Source & Tools

Memorilla compresses document collections into 16 memory tokens for frozen LLMs

Stanford researchers released Memorilla, an open-source memory system that compresses a document collection into 16 memory tokens for a frozen Qwen3-8B. It achieves 84.33% on FactKG while using 108 times fewer document tokens than full context.

MarkTechPost
Nov 14
Industry & Policy

Amazon pursues World Product Graph to unify global product data

Amazon is developing the World Product Graph, a project to consolidate product information from across the globe, including items it does not sell. The company aims to reduce consumers' need to visit other sites and steer external AI agents to Amazon. It plans to combine brand-supplied data with information from other retailers to identify product differences and relationships.

Korean outletAI Times
Nov 13
Agents

Shield automates legacy portal work for IT technicians using OpenAI

Shield uses OpenAI to help IT technicians hand off work in legacy portals. Its Forge agents navigate systems so technicians can focus on issues requiring their judgment.

OpenAI News
Nov 12
Physical AI

FamilyMart to expand beverage restocking robot nationwide from January

Japanese convenience store chain FamilyMart will begin full rollout of a robot that automatically restocks beverages on shelves across all stores from January. The robot is based on TX SCARA technology developed by Teleexistence.

Korean outletiRobotNews