具身性与默会表征:人工智能能走多远?

Embodiment and Tacit Representation: How Far Can Artificial Intelligence Reach?

人类智能的探究和实现问题既是一个哲学问题,也是一个科学问题。人类智能从本质上说是具身的,但同时也紧密地依赖于计算和表征的大脑。导致人工智能研究困境的根源并非在于它采用了计算—表征的框架,而在于它没能将关于身体的信息涵盖进它对智能程序的描述中去。因此,承载着大量实践性知识的默会表征对未来人工智能研究在拓展解决问题常规以及降低计算复杂性等方面有着重要意义。

The research and the realization of human intelligence is a philosophical inquire and a scientific issue as well.Human intelligence is essentially embodied,but also heavily relies on our computational and representational brain.The difficulties artificial intelligence face nowadays,are not resulted from its framework of computation-representation,but rather,the absence of bodily information in the description of intelligence program.Therefore,the tacit representations which bear enormous knowing-how are extremely meaningful to the future research of artificial intelligence,in terms of extending the routine of problem-solving and reducing the complexity of computation.