摘要
随着计算机科学和互联网技术的发展,人工智能进步迅速。科学家和工程师已经制造出在特定认知任务或智力游戏方面能战胜人类的计算机程序。在赞叹的同时,有些人为计算机的强大能力感到担心,怕的是未来人类将要被超级聪明的机器统治;另一些人则对计算机的能力不以为然,认为类似只会下围棋的AlphaGo远谈不上聪明。要讨论人工智能是不是比人类聪明,首先要区分人工智能的含义,进而论证在狭义上,计算机确实比人聪明,最后初步探讨人工智能在通用的意义上,甚至最强的意义上是否可以达到(甚至超过)人类的水平。
Scientists and engineers have created computer programs that can beat humans at specific cognitive tasks or mind games. While marveling at the power of computers, some people are worried about their capabilities, fearing that in the future human beings, especially those unskilled clerks and editors, will be repalced and even ruled by super-smart machines;others are unconvinced of the computers’ abilities, believing that AlphaGo is far from smart. Intelligence has various meanings, and therefore there are strong AI and weak AI, while within weak AI,there are narrow intelligence and general intelligence. I argue that in the narrow sense, computers are indeed more intelligent than human;but not in any stronger sense. I also explore the question whether AI can reach(or even surpass) humans in a generic sense, or even in the most strong sense of intelligence, and my conclusion is largely affirmative.
作者
李楠
LI Nan(Gennlife Technology co.,Ltd Beijing 100080,China)
出处
《科学.经济.社会》
2021年第1期39-46,共8页
Science Economy Society