Smart machines learn by playing games, making mistakes and learning from them. To solve new challenges not stored in past experience, machines create new paths through data. Machines learn quickly, forge their way in unanticipated directions through the neural network frameworks constructed by Google, Facebook, Amazon, etc. This burrowing through networks of neuron-aping nodes, called deep-learning sometimes results in unpredictable behavior. If machines behave in unexpected ways, their human creators have lost control.
Organizations have formed to develop ethics standards for robotics and AI. Refer to Open Roboethics and the Foundation for Responsible Robotics to watch the progress of Robot ethics standards development.
Dr Jason Millar, The Guardian, March, 2017 “The momentous advance in artificial intelligence demands a new set of ethics”