Uploaded November 2026 | Updated November 2026, 3 days ago
Gemini 4 Carbon is Google's new model for long engineering tasks. I had early access, so this video skips the benchmarks and shows the outputs: a Three.js open world, a 3D studio, a house built from a floor plan, a historic fort rebuilt over 100+ unattended passes in a one-month-old engine, GPU experiments, and a neural network written by hand with no training. You'll also get the four prompting rules that made the biggest difference.
Google's announcement: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/
67.js, the 3D engine used for the fort and the jet engine: github.com/codepawl/67.js
The Three.js post the runner demo recreates: https://x.com/threejs/status/2119548992346533900
My video on weight-space programming: youtu.be/wXJDtImT_-0
Credits: dragon concept art by Inigo Quilez (CC0). House drawings: Historic American Buildings Survey, Library of Congress. Fort reference photos: NoahOmarY (CC BY-SA 4.0), Ibrahim Zauq (CC BY-SA 3.0), Marwatt (public domain), via Wikimedia Commons. Satellite image: Esri World Imagery.
Note: the fort footage labelled "pass 101" was rendered after I recorded the narration, which describes the run at pass 65.
Disclosure: I recently joined Google, and I started this testing before that. Google DeepMind gave me early access to this model. These are my own views.
00:00 - Intro: Gemini 4 Carbon & Beyond Benchmarks
02:10 - Three.js Browser Demo & Self-Auditing
03:21 - Math-Only 3D Modeling (SDF Dragon)
04:26 - Sub-Agents: "Storm in a Teacup" Visualization
05:28 - Interactive YouTube Studio Planner
06:45 - 1949 Blueprint to 3D Walkthrough
07:29 - 20-Hour Unattended Fort Reconstruction
10:12 - Autonomous Review Loops & Sub-Agent Scoring
11:03 - Prompting Rules for Long-Horizon Tasks
12:25 - Hardware Physics: Colliding Galaxies Simulation
13:06 - Reinforcement Learning Quadruped Robot
13:49 - Zero-Shot Neural Network Weights (Breakout)
15:52 - Key Takeaways for Agentic Workflows
17:09 - Final Fort Progress & Conclusion
Gemini 4 Carbon is Google's new model for long engineering tasks. I had early access, so this video skips the benchmarks and shows the outputs: a Three.js open world, a 3D studio, a house built from a floor plan, a historic fort rebuilt over 100+ unattended passes in a one-month-old engine, GPU experiments, and a neural network written by hand with no training. You'll also get the four prompting rules that made the biggest difference.
Google's announcement: https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-carbon/
67.js, the 3D engine used for the fort and the jet engine: github.com/codepawl/67.js
The Three.js post the runner demo recreates: https://x.com/threejs/status/2119548992346533900
My video on weight-space programming: youtu.be/wXJDtImT_-0
Credits: dragon concept art by Inigo Quilez (CC0). House drawings: Historic American Buildings Survey, Library of Congress. Fort reference photos: NoahOmarY (CC BY-SA 4.0), Ibrahim Zauq (CC BY-SA 3.0), Marwatt (public domain), via Wikimedia Commons. Satellite image: Esri World Imagery.
Note: the fort footage labelled "pass 101" was rendered after I recorded the narration, which describes the run at pass 65.
Disclosure: I recently joined Google, and I started this testing before that. Google DeepMind gave me early access to this model. These are my own views.
00:00 - Intro: Gemini 4 Carbon & Beyond Benchmarks
02:10 - Three.js Browser Demo & Self-Auditing
03:21 - Math-Only 3D Modeling (SDF Dragon)
04:26 - Sub-Agents: "Storm in a Teacup" Visualization
05:28 - Interactive YouTube Studio Planner
06:45 - 1949 Blueprint to 3D Walkthrough
07:29 - 20-Hour Unattended Fort Reconstruction
10:12 - Autonomous Review Loops & Sub-Agent Scoring
11:03 - Prompting Rules for Long-Horizon Tasks
12:25 - Hardware Physics: Colliding Galaxies Simulation
13:06 - Reinforcement Learning Quadruped Robot
13:49 - Zero-Shot Neural Network Weights (Breakout)
15:52 - Key Takeaways for Agentic Workflows
17:09 - Final Fort Progress & Conclusion










