【作者】 罗诚；

【导师】 任佩瑜；

【作者基本信息】 四川大学 ， 企业管理， 2021， 博士

【摘要】 在社会历史发展过程中,颠覆性技术一直是社会变革和社会生产力发展的革命性力量。改良的蒸汽机和珍妮纺织机的出现,推动了第一次工业革命,使人类社会从农业社会迈向了工业社会;当汽油机和电力的出现,又使人类社会发生了第二次工业革命,社会生产力和人类文明高速发展;二十世纪六十年代,计算机技术和信息技术的发展,使人类社会发生着第三次革命,社会发生了翻天覆地的变化,我们进入了信息化时代。颠覆性技术的出现不仅能够改变人们的生活和工作方式,还能使我们享受它给社会带来的便利和红利。近年来,随着高科技的发展和产业结构变革的快速演进,重大的颠覆性技术持续不断地涌现,推动了新业态、新产品和新需求的产生,影响着经济格局和产业形态的调整,并且越来越成为驱动社会经济发展和提高国家竞争力的关键因素。本文所研究的基于深度学习的人工智能颠覆性技术,将会是在互联网的流量红利逐渐消失以及互联网+的机会越来越有限的情况下,成为一项真正能够提高社会生产力,解决供需失衡的颠覆性技术。人工智能这项颠覆性技术目前在医疗健康、智能安防、金融、制造业等各行业的逐步应用,都在改变着传统的产业格局和人们的生活模式。在供给侧结构性改革的背景下,人工智能颠覆性技术作为一项新的创新事物,必将对市场结构、需求特征、文化创新和社会变革带来重大的影响。但是,颠覆性技术因其革新性及前瞻性,往往需要历经一定的发展过程才能最终被社会所接纳,并进一步得到良好的应用与发展,最终实现商业化。例如,数字摄影技术对传统胶片企业的颠覆、新能源汽车技术对传统汽车技术的颠覆、共享经济对传统经济的颠覆等等,其商业化的历程至少都经历了若干年时间才最终得以实现。在颠覆性技术实现商业化的进程中,整个产业链上的利益相关方对该技术的认知度、接受度、应用度、商业化战略以及外界环境对技术本身以及应用的影响,都会在极大程度上影响颠覆性技术的商业化进程。本文以基于深度学习的人工智能颠覆性技术的商业化作为研究对象,围绕商业化的两个重要阶段即产品研发和市场进入阶段,通过两个完全不同的视角展开研究。一方面在产品研发阶段从用户角度特别是个人消费者角度研究了人工智能颠覆性技术的接受度,得到了一个人工智能颠覆性技术的接受假设模型Ai TAM,通过基于结构方程的实证分析确定并验证了影响人工智能颠覆性技术接受度的影响因素,修正假设模型并得到了最终的理论模型,并分析总结了促进以及妨碍用户接受深度学习的人工智能颠覆性技术的主要因素,针对性地提出了相应的建议和措施,以指导企业在颠覆性技术商业化的产品研发阶段实施正确合理的研发策略;另一方面,在颠覆性技术商业化的市场进入阶段,从企业角度研究了深度学习的人工智能颠覆性技术的商业化战略选择问题,分析了影响颠覆性技术商业化战略的主要因素,得到了一个人工智能颠覆性技术的商业化战略选择模型,并为理论模型的应用建立了相应的三级评价指标体系,企业可以通过计算内外部评价指标的直觉模糊数评价值来得到商业化战略选项,之后通过实证分析以及案例研究,验证了该模型的有效性和可靠性。本文结合了心理学、行为科学、管理学以及统计学等多学科的研究方法,以用户对人工智能颠覆性技术的接受度和企业的人工智能颠覆性技术商业化战略制定为导向,将看似两个独立的领域在深层次上进行交叉融合,解决行为科学与战略管理两个领域中的共性科学问题,即如何快速成功地实现人工智能颠覆性技术的商业化。本文取得的研究成果主要包括以下几个方面:1.从用户角度研究人工智能颠覆性技术接受度时,通过以技术接受模型(TAM)为基础,对理性行为理论(TRA)、创新扩散理论(IDT)和感知风险理论(PRT)等理论进行整合,构建了基于深度学习的人工智能技术接受模型(Ai TAM),并通过实证分析的结果和研究讨论,提出相应的措施和建议以解决人工智能颠覆性技术在产品研发阶段的障碍。该模型综合考量了心理、风险、行为规范等各个领域对人工智能颠覆性技术接受度的影响因素,全面合理地解释了该技术在商业化过程中产品研发阶段用户接受度的影响要素,填补了理论研究的空白。2.从企业的角度来研究人工智能颠覆性技术的商业化战略选择时,首先根据大量的文献研究工作,构建了一个以企业内部综合能力为横轴、外部环境不确定性影响度为纵轴的二维颠覆性技术商业化战略选择模型,提出了技术许可战略、横向平台合作战略、纵向平台协作战略、商业竞争战略四种可选择的商业化战略并在二维平面上给出了四种战略的覆盖区域。该模型的应用采用了基于直觉模糊数评价的分析方法,首先创新性地建立了由25项评价指标构成的三级评价指标体系,通过实证数据验证了指标的独立性及合理性,并设计了相应的量表;之后在此评价指标基础上利用直觉模糊数评价法计算研究企业的内部综合能力以及外部环境不确定性影响度的综合评估值,根据这两个值可以确定企业的商业化战略选择点在理论模型中的位置,得出企业在当前条件下应当采取的商业化战略选项。3.为验证基于深度学习的人工智能颠覆性技术商业化战略选择模型的有效性及可靠性,本文采用了定性的案例研究的的方法,本文选择了以川大智胜、海康威视、大华股份和歌华网络等七家拥有基于深度学习的人工智能技术的代表性企业作为研究对象,通过高管的深度访谈、实地调研以及数据挖掘上市公司数据等方法得到了企业针对深度学习的人工智能颠覆性技术所实际采取的商业化战略,并与本文通过理论模型计算得到的结果进行了对比分析,对比结果显示本研究的理论模型是有效并且可靠的。4.根据得到的两个理论模型以及相关研究结果,本文最后根据用户对人工智能颠覆性技术的接受度在产品开发阶段提出了相关的建议和措施,在市场进入与开发阶段对企业的相关商业化战略选择以及政府在制定相关的政策法规以及产业规制时提出了一些有针对性的建议和措施,以期为人工智能颠覆性技术的快速商业化做出一定的贡献。更多还原

【Abstract】 In the process of social and historical development,disruptive technology has always been a revolutionary force for social changes ~1 and social productivity development.The emergence of improved steam engines and Jenny textile machines promoted the first industrial revolution and brought human society from an agricultural society to an industrial society.When gasoline engines and electricity emerged,the second industrial revolution occurred in human society.Productivity and human civilization developed at a high speed;in the 1960s,the development of computer technology and information technology caused a third revolution in human society,which caused earth-shaking changes in our society.We entered the information age.The emergence of disruptive technology can not only change the way people live and work,but also enable us to enjoy the convenience and dividends it brings to our society.In recent years,with the rapid development of high-tech and industrial structure reform,major disruptive technologies have emerged continuously,which has promoted the emergence of new formats,new products and new demands,and affected the adjustment of economic structure and industrial form.And it is increasingly becoming a key factor driving social and economic development and improving national competitiveness.As the artificial intelligence disruptive technology studied in this paper,when the traffic of the Internet has disappeared and the opportunities of Internet+are very limited,artificial intelligence disruptive technology based on deep learning will become a subversive technology that can truly improve social productivity and solve the imbalance between supply and demand.For example,the gradual application of artificial intelligence disruptive technology based on deep learning in medical health,intelligent security,finance,manufacturing and other industries is changing the traditional industrial pattern and people’s life mode.In the context of supply-side structural reform,disruptive technology,as a new innovation,will have a significant impact on market structure,demand characteristics,cultural innovation and social change.However,because of its innovation and foresight,disruptive technology often needs to go through a certain development process to be finally accepted by the society,further applied and developed,and finally commercialized.For example,the digital photography technology subverts the traditional film enterprises,the new energy vehicle technology subverts the traditional automobile technology,and the sharing economy subverts the traditional economy.Its commercialization process has taken at least several years to finally realize.In the process of commercialization of disruptive technology,the awareness,acceptance,application,commercialization strategy of stakeholders in the whole industrial chain and the impact of the external environment on the technology itself and application will greatly affect the commercialization process of disruptive technology.This paper takes the commercialization of AI disruptive technology based on deep learning as the research object,and focuses on the two important stages of commercialization,namely product R&D and market entry,through two completely different perspectives.On the one hand,in the product R&D stage,the acceptance of AI disruptive technology is studied from the perspective of users,especially individual consumers,and an acceptance hypothesis model aitam of AI disruptive technology is obtained.Through empirical analysis based on structural equation,the influencing factors affecting the acceptance of AI disruptive technology are determined and verified,Revise the hypothetical model and obtain the final theoretical model,analyze and summarize the main factors that promote and hinder users from accepting AI disruptive technology for in-depth learning,and put forward corresponding suggestions and measures to guide enterprises to implement correct and reasonable R&D strategies in the product R&D stage of commercialization of disruptive technology;On the other hand,in the market entry stage of the commercialization of disruptive technology,this paper studies the commercialization strategy selection of AI disruptive technology with deep learning from the perspective of enterprises,analyzes the main factors affecting the commercialization strategy of disruptive technology,and obtains a commercialization strategy selection model of AI disruptive technology,The corresponding three-level evaluation index system is established for the application of the theoretical model.Enterprises can obtain the commercialization strategy option by calculating the intuitive fuzzy number evaluation value of internal and external evaluation indexes.Then,the effectiveness and reliability of the model are verified through empirical analysis and case study.This paper combines multi-disciplinary research methods such as psychology,behavioral science,management and statistics.Guided by the user’s acceptance of AI disruptive technology and the formulation of the enterprise’s AI disruptive technology commercialization strategy,this paper integrates the seemingly two independent fields at a deep level,Solve the common scientific problems in the two fields of behavioral science and strategic management,that is,how to quickly and successfully commercialize the disruptive technology of artificial intelligence.The research results obtained in this paper mainly include the following aspects:1.When studying AI disruptive technology acceptance from the perspective of users,based on the technology acceptance model(TAM),this paper integrates the theories of rational behavior theory(TRA),innovation diffusion theory(IDT)and perceived risk theory(PRT),constructs an AI technology acceptance model based on deep learning(aitam),and discusses the results of empirical analysis and research,Put forward corresponding measures and suggestions to solve the obstacles of AI disruptive technology in product R&D stage.The model comprehensively considers the influencing factors of psychology,risk,code of conduct and other fields on the acceptance of AI disruptive technology,comprehensively and reasonably explains the influencing factors of user acceptance in the product R&D stage of the technology in the commercialization process,and fills the gap of theoretical research.2.When studying the commercialization strategy choice of AI disruptive technology from the perspective of enterprises,firstly,according to a large number of literature research work,this paper constructs a two-dimensional subversive technology commercialization strategy choice model with the enterprise’s internal comprehensive ability as the horizontal axis and the external environment uncertainty as the vertical axis,and puts forward the technology licensing strategy,horizontal platform cooperation strategy Vertical platform cooperation strategy and business competition strategy are four optional commercialization strategies,and the coverage areas of the four strategies are given on the two-dimensional plane.The application of the model adopts the analysis method based on intuitionistic fuzzy number evaluation.Firstly,a three-level evaluation index system composed of 25 evaluation indexes is innovatively established,the independence and rationality of the indexes are verified by empirical data,and the corresponding scale is designed;Then,on the basis of this evaluation index,the intuitionistic fuzzy number evaluation method is used to calculate and study the comprehensive evaluation value of the enterprise’s internal comprehensive ability and the uncertainty of the external environment.According to these two values,the position of the enterprise’s commercialization strategy selection point in the theoretical model can be determined,and the commercialization strategy options that the enterprise should adopt under the current conditions can be obtained.3.In order to verify the effectiveness and reliability of the commercialization strategy selection model of AI disruptive technology based on deep learning,this paper adopts the method of qualitative case study.This paper selects seven representative enterprises with AI technology based on deep learning,such as Sichuan University Zhisheng,Hikvision,Dahua shares and Gehua network,as the research object,Through in-depth interviews with executives,field research and data mining of data of listed companies,this paper obtains the actual commercialization strategy adopted by enterprises for the artificial intelligence subversive technology of in-depth learning,and makes a score analysis with the results calculated by the theoretical model in this paper.The comparison results show that the theoretical model in this study is effective and reliable.4.According to the two theoretical models and relevant research results,this paper finally puts forward relevant suggestions and measures in the product development stage according to the user’s acceptance of AI disruptive technology,In the market entry and development stage,it puts forward some targeted suggestions and measures for the relevant commercialization strategy choice of enterprises and the formulation of relevant policies,regulations and industrial regulations by the government,in order to make a certain contribution to the rapid commercialization of artificial intelligence disruptive technology.更多还原