I built an AI Tamagotchi before Meta could
Sean Hollister from The Verge built a $50 custom AI device using Waveshare ESP32-S3 instead of buying Meta's upcoming Muse Charm.
Stock photo for illustration only, not from the actual event
- The Verge senior editor Sean Hollister built his own AI charm for just $50 instead of buying Meta's Muse Charm.
- The device utilizes a Waveshare ESP32-S3 1.75C devkit combined with custom Claude-generated code.
- Features include ElevenLabs AI voice output and Bluetooth remote camera shutter controls.
- The custom software introduces a Tamagotchi-style minigame where users must feed and care for avatars.
Meta CEO Mark Zuckerberg captured tech headlines by revealing the Muse Charm, a squishy and cute wearable device designed to let users carry their AI assistant around like a digital Tamagotchi ahead of the holiday season.
However, rather than purchasing the official accessory, The Verge senior editor Sean Hollister decided to assemble his own version for a mere $50, noting that the hands-on learning experience was well worth the effort. While his DIY gadget lacks built-in cameras or 5G cellular connectivity, it successfully communicates with Muse, displays adorable avatars, functions as a pocket watch, and theoretically controls smart home appliances.
Stock photo for illustration only, not from the actual event
The compact hardware used for this project is the Waveshare ESP32-S3 1.75C, a 1.75-inch touchscreen computer pendant shipped directly from China. This push-to-talk AI devkit is officially sanctioned by Meta as a suggested platform for developers looking to build devices that connect directly to the Muse AI cloud using open-source code.
Upon unboxing the silver metal pocket watch, Hollister initially found it largely impractical. Although the basic setup involved simply connecting the device to a PC and pulling repository files from Meta's GitHub, the system initially mistook his commands for audio transcription clips, suffered from slow response times, and lacked audio output.
"My version doesn't sport cameras or a 5G cellular radio, but it really talks to Muse and has cute avatars, too!"
Sean Hollister
The methodology demonstrated in this project highlights the rising trend of Vibe Coding, where enthusiasts leverage advanced AI models like Claude to write and modify complex software implementations through natural language prompts, bypassing traditional and tedious coding barriers.
To overcome the hardware's initial limitations, Hollister utilized Claude Code to program conversation history screens, integrate a British AI voice output from ElevenLabs, enable low-latency Bluetooth camera previews, and develop a Tamagotchi-inspired feeding and care minigame for the digital avatars.
Source: The Verge
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment