Harnessing the Power of Neural Operators with Automatically Encoded Conservation Laws