MIT presents a way to explain autonomous-driving AI decisions
· By Antonio Sedino, CTRO · Published by Reinventy Solutions Corp.
CW-Net translates an autonomous vehicle’s AI reasoning into concepts people can understand.
MIT News describes a method called CW-Net that translates the reasoning process of an autonomous vehicle’s AI system into understandable concepts. The work is presented as helping people anticipate when a self-driving car might make a mistake. The available account does not establish a guarantee of error-free driving.
Read the original source at MIT News — Artificial Intelligence ↗