When watching the Neuro/Evil IRL stream, I thought it would be nice if you can have an AI guide seamlessly integrated in to your experience when traveling around. It's so happen that I got an AR glasses from my friend, so I wrote an app to implement this.
https://github.com/ECHO-HELLO-WORLD424/NeuroGlasses/tree/main
When implementing this I realized that maybe AR glasses can be used in IRL streams too. Here are some of the ideas that are possible to implement:
-
HUD Can Sync Twitch View (sort of)
During Neuro/Evil's tour, it seemed that Rob need to watch Neuro/Evil/Chat from phone. The glasses can accepts OpenAI API style streaming text/audio input and play them with the glasses' HUD/speaker, which means it should be able to sync textual content on Twitch. While I can't play video on the glasses, It's possible gather different static Neuro images (happy face, cry face, question mark face ...) and Neuro can update the image on HUD via NeuroSDK -
Neuro Can Operate HUD
HUD layout is defined by a simple json string and can be update with a short json string. This enables Neuro to draw whatever she want to draw (e.g., she can draw a box on the item she wants to buy) on the glasses. Meanwhile the HUD (in theory) can be synced to Twitch because it is possible to render the same UI on regular android machines via glasses' SDK