AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Using Theory of Mine in Robotics
Theory of mine is a way to obtain that link and be able to infer the stuff you can't see from what you can. It's easy for these models to pick up on sort of silly patterns that help them explain human behavior, without actually being causal. And so how do i kind of tell the robot, this is not what areal human would ever do? Versus, this is something that's actually plausible. The more you can kind of help the robot by in forming it about that structure, the more in the easier time it can have to make sense of the behavior they're trying to predict.