We should be able to see what LAM is ddoing like on device, and then iff it does the action as asked, then give it a carrot. it helps with transparency at the very least. But if you deployed it like this, Boom! scalable and helpful reinforcement learning
4 Likes
Love the idea!
1 Like
Great idea but the logistics get iffy. How are you going to live transmit the screen to your small r1 screen without using up people’s expensive data plans? lol
1 Like
after the actual action is done on the rabbit servers, a simplified version gets sent to the device 
1 Like