像教育人一样教育机器——人类教学原则能用于通用人工智能系统吗

Aiming to explore the general laws of intelligence,artificial general intelligence( AGI) has many human-like cognitive functions such as perception,memory,emotion,reasoning,decision-making,etc. This article takes the AGI System as a machine learner to investigate whether the pedagogical effect of human teaching principles could be found in the machine learners. The results of controlled experiments show that changes,including teaching time,teaching content,and teaching methods,lead to significant changes in the network structure of memory of the AGI System. It not only confirms the possibility of transferring human teaching principles to the AGI System,but also reflects the existence of universal and falsifiable educational rules beyond human and machine. In the high-tech era,Artificial Intelligence( AI) has become a key external source of innovation,but the internal source in education is often in a weak position. However,AGI expands the research objects of pedagogy,and broadens the boundaries of pedagogical