人工智能将走向哪里:思维机器会出现吗?

Where Will Artificial Intelligence Go:Will Thinking Machines Emerge?

图灵提出“机器可以具有智能”的观点受到了各种批判,其中有关“机器能否具有意识”的观点引起了学界讨论,塞尔通过“中文屋”思想实验与图灵测试进行类比,指出机器没有智能的原因在于缺少产生意向性的生物基础;德雷福斯把人工智能从技术领域延伸到哲学视角,分别从生物学、心理学、认识论和本体论4个角度批判人工智能建构的基础,重新审视建构思维机器的象征性方法,并指出抽象符号不能完全等同于现实世界。制造思维机器还面临默会限制、人为限制、伦理限制等现实挑战,按照珀尔的观点,奇点时代的到来还需要攻克计算机内部运行程序的准确性和透明性、找到获得自由意志的基础以及整合机器内部单一算法等难关。

Turing’s vision that“machines can have intelligence”has been a controversial issue,among which the view of“whether machines can have consciousness”has aroused a heated debate in a number of research.Searle compared Turing Test to a thought experiment of“Chinese Room”,pointing out that the reason why machines have no intelligence lies in its lack of biological basis for intentionality.Dreyfus extends Artificial Intelligence from technology to philosophy,criticizes the basis of the construction of Artificial Intelligence from the perspectives of biology,psychology,epistemology and ontology,and re-examines the symbolic methods of constructing thinking machines,pointing out that abstract symbols cannot be equated completely with the real world.The manufacture of thinking machines also faces practical challenges such as default,artificial,and ethical restrictions.According to Pearl,the arrival of the singularity era also needs to overcome such difficulties as accuracy and transparency in running programs inside the computer,identifying the basis for obtaining free will,and integrating a single algorithm within the machine.