人工智能时代的计算主义世界观与计算智能限度

The Computationalist Worldview and the Limit of Computational Intelligence in the Age of Artificial Intelligence

  • 摘要: 研究从计算主义世界观的7个根本问题入手,结合图灵、哥德尔、王浩和当代量子计算科学家的相关思考,重新探讨了人工智能或计算智能的本体论限度、世界的(可)计算性等问题,并尝试对三者之间的某些深层关系进行勾勒。这些问题都指向了对人类认知和人工智能认知的局限性与不确定性,以及对智能-世界关系的未完成探求。物理宇宙和抽象数学宇宙都存在心灵不可触及的界限和盲区,即使承认改进版的泛计算主义,宇宙具有的计算性的特征也不意味着其中的一切都对于人的心灵是可计算的。

     

    Abstract: Starting from seven fundamental questions within the computationalist worldview, the ontological limits of artificial intelligence or computational intelligence, as well as the computability of the world, are reexamined by integrating insights from Turing, Gödel, Hao Wang, and some contemporary quantum computing theorists. Some deep interconnections among these three domains are outlined. These issues all point toward the unfinished research regarding the limitations and uncertainties inherent in human and artificial cognition, as well as the relationship between intelligence and the world. Both the physical universe and the abstract mathematical universe contain boundaries and blind spots inaccessible to human mind. Even if an modified pancomputationalism can be accepted, the computational characteristics of the universe does not imply that everything within it is computable for human mind.

     

/

返回文章
返回