Imagine a device that understands everything you say, responds intelligently to your questions, and adapts to your needs… That’s exactly what Project 01 offers, an open-source framework that combines artificial intelligence and a voice interface.
Inspired by the iconic Star Trek computer and developed by the same people who created the beloved Open Interpreter, 01 aims to become the GNU/Linux of voice assistants. Its open and modular ecosystem allows developers to create innovative applications and customize the user experience. Whether you’re a maker or a curious novice, you can build your own 01 device by following the detailed instructions provided on the GitHub repository.
The 01 Light model, based on an ESP32, is an excellent starting point for discovering this fascinating technology. Naturally, everything is open-source, and you can create your own device or run it on an existing one.
For an optimal experience, 01 works in conjunction with a server running on your personal computer or NAS. This server uses tools like Whisper for voice recognition and GPT for generating responses, enabling natural and relevant interactions.
This is made possible thanks to the LMC (Language Model Computer) format, inspired by Andrej Karpathy’s concept of “LLM OS.” 01 uses a language model that interprets code to respond to system events. LMC messages extend the OpenAI format by adding a “computer” role that executes Python code directly within system instructions.
Its flexible architecture allows it to be integrated into a multitude of everyday objects, from voice assistants to household robots and smart cars. The possibilities are endless and just waiting to be explored.
As an open-source project, 01 relies on a dynamic community of contributors to develop and improve. There’s even a Discord server to connect with other project enthusiasts.
