# Functional smart devices using Gemini 4 Argon and Google Antigravity

## Executive summary

This video demonstrates how the makers used Gemini 4 and Google Antigravity to rapidly prototype a functional 'Surf Clock'—a physical device that displays local surf conditions. The process showcased an end-to-end, AI-assisted workflow, moving from initial conceptual sketches (targeting a retro 60s design) to generating 3D models, writing custom firmware, and even producing an assembly video guide. The project was completed in three days with minimal material cost.

## Key takeaways

- AI-Driven Design Iteration: Gemini 4 was used to explore multiple design concepts and refine the physical object, focusing on simplicity and modularity (e.g., snap-together parts and extruding letters to eliminate painting).
- Automated 3D Modeling and Engineering: The AI utilized tools like Blender, Python, and OpenSCAD to generate the complete 3D files for the device, including housing areas for easy assembly.
- Full-Stack Functionality: The process extended beyond physical design, involving the writing of custom software and firmware to pull live ocean data and manage complex concepts like bathymetry.
- Assembly and Documentation: Gemini 4 was capable of generating a detailed, instructional video guide for the final assembly of the complex hardware.

## Technical details

- AI Tools and Platforms: Gemini 4 and Google Antigravity were the primary tools used to guide the entire build process, from conceptual design to final assembly instructions.
- Modeling and Scripting: The 3D modeling phase leveraged a mix of professional tools, including Blender, Python, and OpenSCAD, with Gemini 4 reportedly controlling these tools programmatically.
- Hardware and Firmware: The project required sourcing specific electronics, including controllers and stepper motors, and writing custom firmware to process live ocean data.

## Practical implications

- AI models can significantly accelerate the hardware prototyping lifecycle, automating tasks traditionally requiring specialized skills (CAD, firmware coding, technical writing).
- The integration of AI across design, software, and documentation suggests a shift toward highly accessible, low-barrier-to-entry physical computing projects.
- The ability to generate detailed assembly guides reduces the need for specialized technical documentation teams.

## Topics

Internet of Things (IoT), Hardware Prototyping, Generative AI, Embedded Systems, Product Design, https://goo.gle/3UmeYWb, https://goo.gle/developers

Source: https://www.youtube.com/watch?v=pQ_Dnr9Q4k8