|
- 2017
计算机模拟大脑与功能性计算策略Keywords: 人工智能, 图灵测试, 反转图灵测试, 功能, 结构, 算子artificial intelligence, Turing Test, Reverse Turing Test, function, structure, operator Abstract: 目前,计算机模拟大脑的研究取得了重大进步。在分析了当前技术成就之后,文章对“计算机模拟大脑”或“人工智能”的基本标准,即图灵测试的逻辑意义提出质疑,并指出,因为缺乏关于“心智”本身的有效认识,计算机模拟心智的努力可能走上了歧途。文章通过“反转图灵测试”的思维实验论证了计算机的计算策略与人类计算策略的区别,提出,计算机采用的是算法计算策略而人类大脑采用的是算子计算策略。文章进一步在本体论的层面上定义了算法策略是结构性计算,而算子策略是功能性计算。计算机要成功模拟人类大脑的工作需要采用功能性计算。功能性计算具有三个特征:第一,功能性计算在逻辑上将功能置于结构之先,用功能来定义结构;第二,功能性计算的算子同时蕴含了时间和空间意义;第三,功能性计算是连续量的计算。Till now,computer simulation of brain has made so many wonderful achievements. Throughreviewing these achievements,this paper is intended to question the criteria of computer simulation or artificialintelligence( i.e. the logical meaning of Turing Test),and to argue that computer simulation of mind mayfollow a wrong approach to the understanding of human mind. Through a“ Reverse Turing Test”,it is foundthat computer and man calculate with different strategies. The computer takes the arithmetic-based calculationand man the operator-based strategy. On the ontological level,we define the arithmetic-based calculation asstructural calculation and operator-based calculation as functional calculation. If computer wants to simulatehuman mind successfully,it should use the functional calculation strategy. The functional calculation has threefeatures:first,giving the function logical precedence over the structure,i.e. defining structure with function;second,the operator of functional calculation containing the meanings of both time and space simultaneously;third,functional calculation dealing with the task characterized by continuity rather than discreteness
|