人工智能是适格的刑事责任主体吗?

Is Artificial Intelligence a Qualified Subject of Criminal Responsibility?

肯定“人工智能是适格的刑事责任主体”的主张,是以未来强人工智能的实现为前提的。但如果将智能理解为不仅是编程算法还包括心智意识,则这一主张是不能成立的。设计具有人类心智的人工智能,在根本上是不可能实现的。首先,在认知方面,人工智能仅在单项领域具有较高的认知能力,而在跨界领域却存在难以逾越的技术瓶颈。其次,在意识方面,人工智能既不可能内生自发地生成意识,也不可能外在强行地输入意识,因而不可能具有自由意志。最后,在情感方面,人工智能不可能具有人类的情感动机,无法体验犯罪之乐和刑罚之苦,因而不是适格的受罚主体。

The proposition that "artificial intelligence(AI) is a qualified subject of criminal responsibility" is based on the realization of strong AI in the future. But if intelligence is understood not only as programming algorithm, but also as mental awareness, then this proposition is untenable. It is fundamentally impossible to design an AI with the same mental consciousness as human beings. Firstly, in terms of cognition, AI has high cognitive ability only in certain single fields, but is faced with insurmountable technical bottlenecks in crossover fields. Secondly, in terms of consciousness, AI can neither form consciousness spontaneously within itself, nor import consciousness externally, so it is impossible for AI to have free will. Finally, in terms of emotion, AI cannot have the same emotional motivation as human beings, or experience the joy of crime and the suffering of punishment, so it is not a proper subject of punishment.