I had every intention of recording a video of this, but it’s been a long day and I’ll have to post tomorrow. I’ve got Neon AI running fully on Mac OS, no containers, on an M1 Apple Silicon chip! I’ve got some custom OpenWakeWord models in play with skills configured to give the assistant just a touch of personality.
There’s also some prototypes I’ve put together of satellite/node devices connecting to a central Neon system running in my house on an old laptop.
I’ve also been experimenting with a Mark 2 image that features custom wallpaper, a custom wakeword, connectivity to a local LLM, and instead of a “beep” to acknowledge the wakeword it returns a sharp “what?” 