# Notes on frontier training data, physical intelligence & the people behind it.

Field notes, benchmarks and guides from the team collecting and grading the data that trains the next generation of AI models.

### [WorldDraft: The Umbrella Opened in the Wrong Colour](/content/blogs/worlddraft-video-models-4d-world-source-leakage/index.html)

WorldDraft asks whether video models can add realism without rewriting an executable 4D world. The umbrella opens correctly—and…  
Nov 15, 2026 · 15 min read

### [Gemini Robotics 2.5 Hit 96%. What Was the Denominator?](/content/blogs/gemini-robotics-2-5-mercedes-96-percent-success/index.html)

DeepMind reports 96% success for Apollo 2 on Mercedes-Benz lines. Here is what the denominator, resets and recovery…  
Nov 15, 2026 · 14 min read

### [One Mug, Three Teachers: Egocentric Video vs Teleoperation vs UMI](/content/blogs/egocentric-video-teleoperation-umi-robot-training-data/index.html)

Egocentric video, teleoperation and UMI-style grippers can record the same task but preserve different evidence. Here is how…  
Nov 15, 2026 · 13 min read

### [AlphaProof Code: The AI Wrote 6.5 Million Lines. Humans Had to Trust 26,000.](/content/blogs/alpha-proof-code-signal-formal-verification/index.html)

AlphaProof Code generated 6.5 million lines of machine-checked proof for Signal. The deeper story is why specification—not code—may…  
Nov 13, 2026 · 13 min read

### [MKA: The Agent Was Given a Photograph. It Turned It Into ASCII.](/content/blogs/multimodal-knowledge-agent-mka-visual-reasoning/index.html)

An AI agent receives a road photograph, converts it into ASCII and misses the pothole. What MKA reveals…  
Nov 13, 2026 · 13 min read

### [The AI Read Every Update. It Still Used the Wrong State.](/content/blogs/ai-agent-state-tracking-statebench/index.html)

An AI reads every warehouse update, cites a real one, and still sends the driver to the wrong…  
Nov 10, 2026 · 15 min read

### [Weight-Space Programming: The Model Knew the Answer. It Couldn’t Stop.](/content/blogs/weight-space-programming-llm-edit-neural-network-weights/index.html)

Weight-space programming lets an LLM edit a neural network's weights without training. What GPT-6 Astra's 0.016% Qwen edit…  
Nov 5, 2026 · 21 min read

### [Can AI Close the Loop Between Discovery and Experiment?](/content/blogs/can-ai-close-the-loop-between-discovery-and-experiment/index.html)

Scientific discovery has always involved a difficult exchange between what researchers think might happen and what actually happens…  
Nov 5, 2026 · 13 min read

### [The Rise of Materials Designed by Machines](/content/blogs/the-rise-of-materials-designed-by-machines/index.html)

What if scientists could describe the properties they need and let artificial intelligence help design the material itself?…  
Nov 4, 2026 · 13 min read

### [When Proteins Become a Search Space for AI](/content/blogs/when-proteins-become-a-search-space-for-ai/index.html)

For billions of years, nature has been building proteins. Some carry oxygen through our blood, others help digest…  
Nov 4, 2026 · 21 min read

### [AI Is Beginning to Discover What Science Hasn’t Found Yet](/content/blogs/ai-is-beginning-to-discover-what-science-hasnt-found-yet/index.html)

Artificial intelligence is moving beyond predicting outcomes and analyzing existing research. It is beginning to help scientists formulate…  
Nov 3, 2026 · 17 min read

### [The Database Agreed. Nobody Told Him the Penalty.](/content/blogs/banking-agent-procedure-final-state-check/index.html)

A banking AI closes a deposit correctly but never tells the customer the penalty. Why final-state grading misses…  
Nov 3, 2026 · 20 min read

### [Hinglish Is Not Two Languages](/content/blogs/hinglish-code-switching-speech-recognition/index.html)

Hinglish is one grammar with two vocabularies. Why code-switched calls break transcription, WER and turn-taking, and what a…  
Nov 3, 2026 · 23 min read

### [Small Enough to Run Locally, Smart Enough to Go Multimodal](/content/blogs/small-enough-to-run-locally-smart-enough-to-go-multimodal/index.html)

Artificial intelligence has spent years moving in one direction. Models have become larger, computing requirements have increased, and…  
Oct 29, 2026 · 16 min read

### [Gemini 4 Carbon Is Built to Break the Cybersecurity Bottleneck](/content/blogs/gemini-4-carbon-is-built-to-break-the-cybersecurity-bottleneck/index.html)

Finding a security vulnerability is one thing. Understanding whether someone can exploit it, figuring out why it exists,…  
Oct 29, 2026 · 21 min read

### [The Rise of Persistent AI Agents](/content/blogs/the-rise-of-persistent-ai-agents/index.html)

Imagine you have a report to finish before tomorrow morning. You need to research several websites, compare the…  
Oct 26, 2026 · 12 min read

### [Dessa Learned to Talk Without Text. What Did It Learn to Understand?](/content/blogs/block-dessa-speech-only-ai-conversation-model/index.html)

Block’s speech-only AI puts the information between words at the center of the conversation. The next test is…  
Oct 24, 2026 · 17 min read

### [The Missing Audit Trail of Artificial Intelligence](/content/blogs/the-missing-audit-trail-of-artificial-intelligence/index.html)

Imagine a researcher asking an AI assistant a straightforward question about a scientific discovery. The model responds confidently,…  
Oct 22, 2026 · 18 min read

You're all caught up — that’s every post.
