摘要
ChatGPT逐渐发展成为日常工作的智能助手(AI assistant)。但“智能助手”一词以“助手”这一界面隐喻限定了人工智能辅助技术与人交互的形式和规范。目前“智能助手”的功能界面常常以对话形式出现,其本质是一种社交场景。本研究认为,“智能助手”的界面发展方向将可能导致一种基于人格社交特征的技术接纳和使用鸿沟。本研究基于准社会交往视角、认知负荷理论和依恋理论,针对453名在校大学生的ChatGPT使用情况开展了问卷调查。研究发现,尽管ChatGPT所带来的准社会交往体验能够直接地正向预测人们持续使用ChatGPT的意愿,但准社会交往也会在不同依恋回避水平的用户中通过认知负荷间接预测使用意愿。对于依恋回避水平高的个体而言,准社会交往会正向预测他们的认知负荷,进而降低其持续使用意愿。相反,对于依恋回避水平低的个体,准社会交往会负向预测他们使用ChatGPT时的认知负荷,间接地促进了持续使用意愿。基于上述发现,本研究提出应该重视人工智能辅助技术的交互界面设计对不同人格特征人群的技术接纳的影响,且应开发多种交互界面以适应不同社交特征人群的使用偏好。同时,应谨慎对待如“智能助手”“人机协作”等学术词汇所隐含的交互界面设计倾向,警惕相关词汇的使用成为一种技术和文化霸权的手段。
ChatGPT has evolved into an artificial intelligent(AI)assistant for everyday work.However,the term“intelligent assistant”embraces the interface metaphor of“assistant”and thus determines the form and the norm of interactions between AI assistive technology and human beings.At present,the user interface of“intelligent assistant”often appears in the form of dialog,which is essentially a social interaction context.This paper argues that the adoption of the interface metaphor of“assistants”may lead to a digital divide in technology acceptance and use based on the social characteristics of personality.Drawing upon the parasocial interaction perspective,cognitive load theory,and attachment theory,this paper surveyed 453 college students on ChatGPT use.It was found that while the parasocial interaction with ChatGPT directly and positively predicted users'continuance intention to use ChatGPT,parasocial interaction also indirectly predicted continuance intention to use through cognitive load among users with varying levels of attachment avoidance.For individuals with high levels of attachment avoidance,parasocial interaction positively predicted cognitive load,which in turn negatively predicted their continued intention to use ChatGPT.In contrast,for individuals with low levels of attachment avoidance,parasocial interaction negatively predicted cognitive load and in turn,positively associated with continued intention to use the technology.The above findings indicate that researchers should pay attention to the impact of the interactive interface design of AI-assisted technologies on the technology acceptance of people with different personality traits.The findings also advocate that designers should develop a variety of interactive interfaces to adapt to the usage preferences of people with different social traits.At the same time,the tendency of interactive interface design implied by academic terms such as“intelligent assistant”and“human-computer collaboration”should be treated with caution,so as to prevent the use of related terms from becoming a means of technological and cultural hegemony.
作者
卢嘉辉
李炳楠
蒲培洋
Jiahui Lu;Bingnan Li;Peiyang Pu(School of New Media and Communication,Tianjin University)
出处
《全球传媒学刊》
CSSCI
2024年第2期25-50,共26页
Global Journal of Media Studies
