人工群体智能的超越性及其困境
|
摘要
人工群体智能的发展向我们展示了某些显著超越人类个体水平的智能形式,但这种超越仍不足以实现通用人工智能的目标。从其基础假设来看,人工群体智能系统以心智的计算架构为核心,混合了符号计算和人工神经网络两大范式。同时,人工群体智能反映了人机共存环境下智能的非还原的层级结构。面对中文屋思想实验的挑战和"人工智能第一步谬误"的批评,人工群体智能依旧难以突破传统范式的理论瓶颈和安全风险。
The development of artificial swarm intelligence shows us some intelligent forms that significantly exceed the level of human individuals,but these advancements are still not enough to achieve the goal of general artificial intelligence. Regarding its basic hypotheses,the artificial swarm intelligence system is based on the mental computing architecture,which is a hybrid of symbolic computation and artificial neural network. At the same time,artificial swarm intelligence reflects the non-reductive hierarchically structure under the coexistence environment of human and machine.Faced with the challenge of the Chinese room argument and the criticism of the "AI first step fallacy",artificial swarm intelligence is still difficult to break through the theoretical bottlenecks and safety risks of the traditional paradigm.
The development of artificial swarm intelligence shows us some intelligent forms that significantly exceed the level of human individuals,but these advancements are still not enough to achieve the goal of general artificial intelligence. Regarding its basic hypotheses,the artificial swarm intelligence system is based on the mental computing architecture,which is a hybrid of symbolic computation and artificial neural network. At the same time,artificial swarm intelligence reflects the non-reductive hierarchically structure under the coexistence environment of human and machine.Faced with the challenge of the Chinese room argument and the criticism of the "AI first step fallacy",artificial swarm intelligence is still difficult to break through the theoretical bottlenecks and safety risks of the traditional paradigm.
引文
(1)Surowiecki,J.The wisdom of crowds:Why the many are smarter than the few and how collective wisdom shapes business.Economies,Societies and Nations,2004,p.296.
(2)《新一代人工智能发展纲要(2017)》http://www.gov.cn/zhengce/content/2017-07/20/content_5211996.htm。
(3)参见Reif,W.E.,Biogeochemistry-Biosphere-Noosphere,The Growth of the Theoretical System of Vladimir Ivanovich Vernadsky(Studienzur Theorie der Biologie,Band 4,2001)一书对智能圈和韦尔纳斯基学术思想的介绍。
(4)Seeley,T.D.,Honeybee Democracy,Princeton University Press,2010,p.46.
(5)Mulgan,G.True Collective Intelligence?A Sketch of a Possible New Field,Philosophy&Technology,2014,27(1),pp.135-136.
(6)Li,W.,Wu,W.J.,Wang,H.M.,Cheng,X.Q.,Chen,H.J.,Zhou,Z.H.,&Ding,R.Crowd Intelligence in AI 2.0 Era,Frontiers of Information Technology&Electronic Engineering,2017,18(1),p.16.
(7)Rosenberg,L.B.Human swarming,a real-time method for parallel distributed intelligence.In Swarm/Human Blended Intelligence Workshop(SHBI),2015,pp.1-7.
(8)参见Hope Reese,How“artificial swarm intelligence”uses people to make better predictions than experts.https://www.techrepublic.com/article/how-artificial-swarm-intelligence-uses-peopleto-make-better-predictions-than-experts/.2016.
(9)这一描述传统人工智能的说法最早出现在John Haugeland的著作Artificial Intelligence:The very Idea(MIT Press,1989)中。
(10)(11)(14)玛格丽特·博登:《AI:人工智能的本质与未来》,孙诗惠译,中国人民大学出版社2017年版,第22、114、28页。
(12)雷·库兹韦尔:《奇点临近》,李庆诚等译,机械工业出版社2011年版,第9页。
(13)Simon,The Sciences of the Artificial(3rd edition),MIT Press,1996,p.23.
(15)参见Minsky,Steps toward Artificial Intelligence,Proceedings of the IRE,1961,49(1),pp.8-30.
(16)Searle,J.R.Minds,Brains and Programs,Behavioral and Brain Sciences,1980,3(3),pp.417-424.
(17)Searle,J.R.Twenty-one years in the Chinese Room,in Preston,J.,&Bishop,M.J.,Views into the Chinese room:New essays on Searle and artificial intelligence,OUP,2002,p.51.
(18)(19)Dreyfus H L.A History of First Step Fallacies,Minds&Machines,2012,22(2),p.87,p.91.
(2)《新一代人工智能发展纲要(2017)》http://www.gov.cn/zhengce/content/2017-07/20/content_5211996.htm。
(3)参见Reif,W.E.,Biogeochemistry-Biosphere-Noosphere,The Growth of the Theoretical System of Vladimir Ivanovich Vernadsky(Studienzur Theorie der Biologie,Band 4,2001)一书对智能圈和韦尔纳斯基学术思想的介绍。
(4)Seeley,T.D.,Honeybee Democracy,Princeton University Press,2010,p.46.
(5)Mulgan,G.True Collective Intelligence?A Sketch of a Possible New Field,Philosophy&Technology,2014,27(1),pp.135-136.
(6)Li,W.,Wu,W.J.,Wang,H.M.,Cheng,X.Q.,Chen,H.J.,Zhou,Z.H.,&Ding,R.Crowd Intelligence in AI 2.0 Era,Frontiers of Information Technology&Electronic Engineering,2017,18(1),p.16.
(7)Rosenberg,L.B.Human swarming,a real-time method for parallel distributed intelligence.In Swarm/Human Blended Intelligence Workshop(SHBI),2015,pp.1-7.
(8)参见Hope Reese,How“artificial swarm intelligence”uses people to make better predictions than experts.https://www.techrepublic.com/article/how-artificial-swarm-intelligence-uses-peopleto-make-better-predictions-than-experts/.2016.
(9)这一描述传统人工智能的说法最早出现在John Haugeland的著作Artificial Intelligence:The very Idea(MIT Press,1989)中。
(10)(11)(14)玛格丽特·博登:《AI:人工智能的本质与未来》,孙诗惠译,中国人民大学出版社2017年版,第22、114、28页。
(12)雷·库兹韦尔:《奇点临近》,李庆诚等译,机械工业出版社2011年版,第9页。
(13)Simon,The Sciences of the Artificial(3rd edition),MIT Press,1996,p.23.
(15)参见Minsky,Steps toward Artificial Intelligence,Proceedings of the IRE,1961,49(1),pp.8-30.
(16)Searle,J.R.Minds,Brains and Programs,Behavioral and Brain Sciences,1980,3(3),pp.417-424.
(17)Searle,J.R.Twenty-one years in the Chinese Room,in Preston,J.,&Bishop,M.J.,Views into the Chinese room:New essays on Searle and artificial intelligence,OUP,2002,p.51.
(18)(19)Dreyfus H L.A History of First Step Fallacies,Minds&Machines,2012,22(2),p.87,p.91.