Safety Intelligence and Legal Machine Language: Do We Need the Three Laws of Robotics?

摘要

By the middle of this century, artificial intelligence and robot technology will no longer be considered science fiction fantasy. At the same time that engineers will be addressing all kinds of technical issues, a combination of engineers, social scientists, legal scholars, and policy makers will be making important decisions regarding robot sociability. In all cases, the topmost concern must be robot safety, since the emphasis for the future will be on human-robot co-existence. In this paper we described a safety intelligence concept that meets three criteria: understanding situations, making decisions, and taking responsibility. The SI concept can also be broken down into two dimensions, the first involving special "third existence" ethical guidelines for robots plus a ban on equipping NGRs with human-based intelligence, and the second involving a mix of third existence designation and legal machine language designed to resolve issues associated with open texture risk. If current reflex control research proves successful, robots will someday be equipped with a safety reflex system to ensure that they avoid dangerous situations, especially those involving humans.