# Building a Custom AI Tamagotchi Using Meta's Open Source Muse Framework A DIY approach to wearable AI hardware using affordable development kits and coding assistants. - **Source:** [The Verge (theverge.com)](https://www.theverge.com) - **Published:** 2026-10-31 - **Topic:** [Software & Engineering](https://brifing.si/c/engineering) - **Original article:** - **On Brifing:** https://brifing.si/s/3397d4bd-a56a-424f-85c2-a516b4e03efc ## DIY Alternative to Muse Charm Instead of purchasing the official Meta Muse Charm, a $50 Waveshare ESP32-S3 1.75C development kit offers a similar hardware experience. This pocket-sized touchscreen pendant serves as a base for building custom AI-powered wearables. By leveraging Meta’s open-source tokens, users can connect directly to the Muse AI cloud without buying proprietary retail hardware. ## Initial Setup and Limitations The device arrived as a silver metal pocket watch but lacked immediate utility. Initial setup involved connecting to a PC and pairing with the Muse phone app via GitHub. However, the default configuration failed to provide audio responses, displayed only truncated text, and lacked essential features like smart home discovery or camera controls. ## Enhancing Functionality via Claude Code Using Claude Code, the device was reprogrammed to overcome initial software gaps. A conversation history page was added for better readability, and an ElevenLabs British AI voice was integrated for audible responses. These modifications transformed the basic pendant into a more interactive device that better serves the user's specific daily requirements. ## Camera and Remote Control The developer added a remote shutter button and a live camera preview feature to the pendant. By utilizing Bluetooth, the device can now trigger the phone’s camera from across the room. These custom scripts resolved issues with upside-down image orientation and enabled high-quality photo capture from creative angles without needing home Wi-Fi. ## Gamification and Tamagotchi Features To increase engagement, the developer programmed custom avatars and a Tamagotchi-style minigame. The system tracks four levels of hunger, boredom, and cleanliness. Treats cause the avatars to get dirty, while excessive play leads to fatigue. This demonstrates how generative AI can be used to create dynamic, state-based interactive experiences on hardware. ## Hardware Optimization and Vibe Coding Refining the user experience required 'vibe coding' to fix input latency. By restricting audio transmission to segments over one second, the device stopped misinterpreting short button presses. While some lag persists due to Meta’s server response times, the hardware now functions as a pocket watch with a working three-hand face. ## Security and Privacy Precautions Given concerns regarding data privacy and potential leaks, the device was isolated on a guest Wi-Fi network. The developer maintains a cautious stance toward Meta’s Muse agent, especially regarding access to home networks or personal email. This highlights the importance of sandboxing experimental AI hardware to protect sensitive user information. ## Future of Programmable AI Wearables The project raises questions about the value proposition of dedicated AI hardware versus general-purpose smartphones. If retail devices like the Muse Charm lack deep programmability, development kits may remain the superior choice for enthusiasts. The true potential of these wearables lies in the ability to customize features beyond the manufacturer's original vision. ## Sources covering this story - https://www.theverge.com/gadgets/1013375/muse-charm-gadgets-preview-waveshare-esp32-claude-vibe-code — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._