Google Antigravity /swarm builds 3D pelican biking demo
Google had a whole lot more fun with Carbon and pelicans on bicycles than I did https://t.co/ZGamnKx...
The notable detail is that critic subagents inspect their own rendered output, turning visual verification into an automated loop rather than a human checkpoint. Developers building agent workflows can study how Blender and Three.js integration enables self-correction in creative coding tasks. Teams evaluating multi-agent coding tools will find the swarm architecture a concrete contrast to single-agent assistants like Cursor or GitHub Copilot.
Google CEO Sundar Pichai demonstrated the /swarm feature in Google Antigravity with a demo that builds a playable 3D game of a pelican riding a bicycle around the Mountain View campus. The demo was shared as a fun test of long-horizon coding capability in the Antigravity environment. Developer Simon Willison commented that Google had more fun with the Carbon and pelican-on-bicycle demos than he did.
The /swarm feature orchestrates builder and critic subagents that run iterative loops across Blender and Three.js. The agents check their own renders during the build process, enabling them to refine the 3D game through multiple rounds of generation and visual verification without human intervention in each step.
The demo positions Antigravity's multi-agent coding environment against competitors focused on single-agent workflows, using a visually verifiable task to showcase self-correcting loops. For developers, it illustrates how long-horizon coding agents can combine 3D content tools with real-time render inspection to deliver finished interactive output.
Google had a whole lot more fun with Carbon and pelicans on bicycles than I did https://t.co/ZGamnKx...
Google had a whole lot more fun with Carbon and pelicans on bicycles than I did x.com/sundarpichai/s… Sundar Pichai @sundarpichai Carbon made a pelican demo so you don't have to 😅 fun test of long-horizon coding and the new /swarm feature in Google @Antigravity . Builder and critic subagents ran iterative loops across Blender and Three.js - checking their own renders to build a playable 3D game of a pelican biking around our Mountain View campus.
This report is categorized under AITOP's "Products" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Google Antigravity, subagents, coding assistants, Three.js.
Source attribution: curated from Simon Willison (@simonw).
