Our Work.
|
Ethics is a set of standards that helps humans decide how we ought to act in various situations. We could say that ethics is all about making choices and providing reasons about what influences these choices.
Industrial robots are programmed to accomplish a focused, predetermined task, and there are no ethical decisions they make themselves. Asimov’s Laws of Robotics would be sufficient to curtail their actions, but more recently created robots with machine learning capabilities continuously need to make decisions based on what they learn. Because of this, a creator has very little influence on their robot's decision making. Asimov’s Laws are not sufficient for machine learning robots so we need ethical filters to be built into the software of the robot. We came up with the idea for our STEM project by reading up on Isaac Asimov’s laws and then wondering how modern machine learning robots would handle real life scenarios based on those laws. The laws fell short. But so did we. We could not come to an agreement about which choices were ethically correct because our ideas were divergent. So we decided to build a framework based on research conducted at Brown University and Santa Clara University. We interviewed experts in the fields of machine learning and software engineering. We came up with the idea of a neural network based ethical framework which provides an API for handling ethical situations. |
