人工智能会是一个要命的问题吗?

Is Artificial Intelligence An Issue of Life and Death?

拟人化的人工智能是一个错误的发展方向,因为欲望、情感和价值观是偏心、歧视和敌对的根据,模仿了人性和人类价值观的人工智能就和人类一样危险。创造性不在于联想和组合功能,而在于在破坏游戏规则的同时建立新的游戏规则,一旦人工智能获得自主建立游戏规则的创造能力就危险了。如果人工智能具备了哥德尔反思能力,就很难控制人工智能了,因此安全的人工智能必须限制在没有反思能力的图灵机水平上。

Personification of artificial intelligence is heading in the wrong direction. Since desire,emotion and value are the foundation of bias,discrimination and rivalry,artificial intelligence endowed with the semblance of human nature and human values will become as dangerous as human beings. Creativity lies not in the associative and integral functions,but in the construction of new game rules where existing ones are destroyed,and the artificial intelligence which acquires the creative ability of autonomous construction of game rules will become truly dangerous. Once equipped with Godellian reflectivity,artificial intelligence will get out of control.Hence,safe artificial intelligence should be frozen on the level of the unreflective Turing machine.