The BESTBOT at Stanford University

Machine ethics researches the morality of semiautonomous and autonomous machines. The School of Business at the University of Applied Sciences and Arts Northwestern Switzerland FHNW realized a project for implementation of a prototype called GOODBOT, a novelty chatbot and a simple moral machine. One of its meta rules was it should not lie unless not lying would hurt the user. It was a stand-alone solution, not linked with other systems and not internet- or web-based. In the LIEBOT project, the mentioned meta rule was reversed. This web-based chatbot, implemented in 2016, could lie systematically. It was an example of a simple immoral machine. A follow-up project in 2018 is going to develop the BESTBOT, considering the restrictions of the GOODBOT and the opportunities of the LIEBOT. The aim is to develop a machine that can detect problems of users of all kinds and can react in an adequate way. It should have textual, auditory and visual capabilities. The paper “From GOODBOT to BESTBOT” describes the preconditions and findings of the GOODBOT project and the results of the LIEBOT project and outlines the subsequent BESTBOT project. A reflection from the perspective of information ethics is included. Oliver Bendel will present his paper in March 2018 at Stanford University (“AI and Society: Ethics, Safety and Trustworthiness in Intelligent Agents”, AAAI 2018 Spring Symposium Series).

Fig.: What will the BESTBOT look like?